Eosinophilic Esophagitis

Updated: Jan 03, 2020
Author: Nina Tatevian, MD, PhD, FCAP; Chief Editor: Nirag C Jhala, MD, MBBS 


Primary eosinophilic esophagitis (EoE) is a clinicopathologic entity that is characterized clinically by symptoms related to esophageal dysfunction and histologically by an eosinophil-rich inflammation that is limited to the esophagus. This disease involves both the proximal and distal esophagus.

In 2007, a multidisciplinary group proposed a consensus definition for the diagnosis of EoE based on the typical clinical presentation and pathologic findings in the absence of other known causes of tissue eosinophilia.[1] This group defined primary EoE as a clinicopathological disease characterized by the following:

  • Symptoms including, but not restricted to, food impaction and dysphagia in adults and feeding intolerance and gastroesophageal reflux disease (GERD) symptoms in children

  • Presence of 15 or more eosinophils per high power field (hpf) on esophageal biopsy

  • Exclusion of other disorders associated with similar clinical, histological, or endoscopic features, especially GERD (use of high-dose proton pump inhibitor [PPI] treatment or normal pH monitoring)

Subsequently, as the understanding of EoE evolved, a phenotypic heterogeneity in disease presentation was observed, such as PPI-responsive esophageal eosinophilia. Furthermore, the practical usefulness of the histological criteria was observed to be limited, as the presence of 15 or more eosinophils/hpf on esophageal biopsy samples could not be validated to successfully discriminate among various causes of esophageal eosinophilia, such as GERD, infections, Crohn disease, and hypersensitivity, among others. All of these factors led to a need to review the 2007 recommendations.

Therefore, in 2011, the multidisciplinary group proposed a new conceptual definition for EoE, according to which “Eosinophilic esophagitis represents a chronic immune/antigen mediated esophageal disease characterized clinically by symptoms related to esophageal dysfunction and histologically by eosinophil-predominant inflammation.”[2, 3]

2013 Guidelines from the American College of Gastroenterology (ACG) define eosinophilic esophagitis as the following[4] :

  • The presence of symptoms related to esophageal dysfunction such as dysphagia, food impaction, chest pain, etc
  • Esophageal mucosa with eosinophil-predominant inflammation, up to 15 eosinophils per high power field
  • Mucosal eosinophils limited to the esophagus and persist after a trial of PPIs
  • Exclusion of secondary causes of esophageal eosinophilia 


Eosinophilic esophagitis (EoE) has been hypothesized to be an atopic inflammatory disease. The etiopathogenesis is hypothesized to be an aberrant immune response to antigenic stimulation.[2] The atopic/allergenic nature of antigen is supported by studies that report a climatic and seasonal variation in presentation.[5, 6, 7, 8, 9] Most of these studies show an increase in new cases during spring or fall and some during summer. A high prevalence of EoE in cold and arid zones in the United states also supports the role of environmental factors in EoE.[7]

The esophagus is unique from the rest of gastrointestinal tract, as eosinophils are not a part of normal histology in esophagus. In other parts of the gastrointestinal tract, eosinophils are a normal component of mucosal infiltrate, even in the absence of disease process. In the late 1990s, murine studies demonstrated that eotaxin, an eosinophil chemoattractant chemokine, plays an important role in homing of eosinophils to the mucosal surface of the gastrointestinal tract.[10] The same group of researchers subsequently reported that interleukin-5 (IL-5), a T-helper 2 (Th2) cytokine, plays an important role in the induction of eosinophil trafficking to the esophagus.[11]

Another study demonstrated that delivery of another Th2 cytokine interleukin 13 (IL-13) to lungs induces EoE by inducing IL-5, eotaxin-1, through signal transducer and activation of a transcription (STAT)–6 dependent mechanism.[12] The results of this study established a link between lung and esophageal inflammation and suggested that direct exposure of esophageal mucosa to an allergen is not necessary for EoE to develop.

An in vitro study involving human esophageal mucosal tissue by Straumann et al[13] demonstrated that EoE induces a selective inflammatory response restricted to the esophagus and includes infiltration of esophageal mucosa by IL-5 expressing T-cells, B-cells, and eosinophils and immunoglobulin E (IgE) expressing mast cells. This landmark study provided scientific evidence for an atopic/allergenic basis of EoE.

In 2006, Blanchard et al[14] demonstrated eotaxin-3 as a critical effector molecule for EoE by gene expression profiling. Later, in an attempt to characterize the cytokine expression in EoE, they reported significantly increased esophageal expression of interleukin 4 (IL-4) and IL-5 mRNA in patients with active EoE. They also provided further evidence in their study that IL-13 and IL-5 associate with eosinophil and eotaxin-3 levels in patients with EoE, thereby indicating the key role of adaptive T(H)2 immunity in regulating eotaxin-3–driven esophageal eosinophilia in the absence of a consistent systemic change in cytokines.

Besides T-helper cells in the pathogenesis of EoE, there is also a role of mixed IgE– and non-IgE–mediated allergic response to food and environmental allergens.[15] This is based on the presence of positive skin prick test results and atopy patch test results to various antigens. In most children, EoE appears to be driven by food antigen exposure, which responds to the elimination of common dietary antigens and recurs upon reintroduction of those food antigens. These children have positive skin test result to a panel of allergens,[5, 16, 17] whereas, in adolescents and adults, aeroallergenic sensitization or atopic dermatitis/eczema is associated with about 60% of cases of EoE.[18]

Eosinophils play an integral role in remodeling of esophageal tissues, which is observed histologically as subepithelial fibrosis and manifests clinically as dysphagia in adults and vomiting in children. Eosinophils cause fibrosis through degranulation and secretion of their granule cationic proteins, particularly major basic protein (MBP) and eosinophil peroxidase (EPO) and elaboration of fibrogenic growth factors such as TGF-β, PDGF-BB, and IL-1β. The association of degranulating eosinophils and deposition of their granule cationic proteins in tissues with pathological fibrosis is a recurrent finding in a broad group of eosinophilic illnesses other than EoE, such as hypereosinophilic syndrome and asthma.[19, 20]

Noel et al have proposed that a genetic predisposition may exist.[21] Few reports of familial clustering of cases support this hypothesis.[22, 23]



The first case of eosinophilic esophagitis (EoE) was reported in 1977 in an adult patient[24] ; however, it was not until the 1990s that it came to be recognized as a distinct clinical entity. In 1993, Attwood et al reported first case series of EoE in 12 adult patients and suggested that this is an entity distinct from GERD.[25] Subsequently, Kelly et alreported a case series of 10 pediatric patients with eosinophilic infiltration of the esophagus whose gastrointestinal symptoms were unresponsive to standard antireflux treatment but improved with and amino acid–based formula diet.6  Since then, there has been an implosion of published literature on EoE.

EoE is believed to occur mainly in developed countries, affects both adults and children, and shows a male predominance.[26]  Most patients are white men.[27] A study that compared the initial presenting symptoms of EoE in males and females reported that women more commonly presented with heartburn, whereas men more commonly presented with dysphagia and food impaction, although no differences in endoscopic differences were noted between the sexes.[28]

There is an increasing trend in newly diagnosed cases per year worldwide.[29, 30, 31] Whether this increase in incidence reflects a heightened awareness of EoE or a true increase is yet undetermined.[32]

In 2005, Straumann and Simon reported an average incidence of 1.438 cases of EoE per 100,000 inhabitants over a 16-year observational period in Olten County, Switzerland.[33] They noted a marked increase in newly diagnosed cases in later years. Since EoE is a chronic condition, this increased incidence reflected an increased prevalence. The authors of this study also suggested that the trend reflects a true increase in incidence and not just increased awareness.

Subsequently, Hruz et al, in another demographic study from Olten county in Switzerland, reported an increased incidence in 6 years from 4.4-7.4 per 100,000 inhabitants per year and a prevalence of 43 patients with EoE per 100,000 inhabitants.[34]  This study in included both the pediatric and adult population. Similarly, in a study from western Australia, an increasing trend in prevalence was observed in the pediatric population over a decade, from 0.05 to 0.89 per 10,000 children.[35]  Ronkainen et al reported a prevalence of 1% EoE in the adult Swedish population.[36]  In New Zealand, the incidence of EoE is 14.1% in esophageal biopsies from people with dysphagia.[30]

In the United States, a pediatric population–based demographic study from Hamilton County in Ohio reported an annual incidence of EoE of 1 per 10,000 and a prevalence of 4.296 cases per 10,000 children.[21] Prasad et al, in a retrospective study on epidemiologic trends of EoE in Olmsted county, Minnesota, reported an increasing incidence of EoE in three decades: 0.86 cases/100,000 population/year from 1976 to 1985 to 8.78 cases/100,000 population/year from 1996-2006.[29]  The prevalence of EoE, in this study, was reported to be 104.7 cases/100,000 population as of January 2007. Sorser et al reported no increase in the incidence of EoE in their cohort of pediatric population over a period of 5 years from 2001-2006.[6]

The current prevalence in developed countries is between 45 and 55 cases per 100,000 population.[37]

An increased prevalence of EoE has been observed in patients with celiac disease.[38, 39] An inverse relationship has been reported between the prevalence of Helicobacter pylori infection and esophageal eosinophilia.[40] Climate has been found to affect EoE prevalence. A recent study by Hurrell et al reported that EoE is most prevalent in the cold and arid zones.[7]


Clinical Features and Imaging

Primary eosinophilic esophagitis (EoE) affects all ethnic groups and both sexes. There is a slight male predominance, with a male-to-female ratio of 3:1.[2] It is a disease of both children and adults.[41]

Generally, the clinical symptoms of EoE are nonspecific, and the patients are in good physical condition; therefore, in some cases, the diagnosis of EoE is made a few years after onset of symptoms.

The presenting symptoms vary depending on the age of onset.[42] Children tend to present with nausea and vomiting, weight loss, anemia, and failure to thrive.[41] In neonates and infants, refusal of food is the most common presenting symptom.[21]

In a 2012 study, Sorser et al[6] reported vomiting as the most common presenting symptom in children and adolescents (61%), followed by dysphagia (39%), abdominal pain (34%), feeding disorders (14%), heartburn (14%), food impaction (7%), vague chest pain (5%), and diarrhea (5%). They further noted that vomiting and feeding disorders affected younger children, whereas heartburn and dysphagia occurred in older children.

In contrast, the characteristic symptom in adults includes dysphagia for solid foods, retrosternal pain, and food impaction.[41] Some patients also present with GERD-like symptoms that are unresponsive to medical or surgical antireflux therapy. However, a subset of patients has been recognized to have a typical clinical presentation of EoE and have had GERD diagnostically excluded, yet show a clinicopathologic response to PPIs. This condition is currently referred to as PPI-responsive EoE.[2]

Other symptoms that often prompt clinical evaluation are throat clearing, choking, gagging, and hoarseness.[41]

Patients in both adult and pediatric age groups often have a concomitant history of allergies such as food allergies, asthma, eczema, or chronic rhinitis. The laboratory workup shows mild peripheral eosinophilia in 5%-50% of children and adults.[37] Elevated total IgE levels are seen in about 70% of patients.[37]

Esophagogastroduodenoscopy (EGD) with biopsy confirms the diagnosis.[41, 43] Radiographic studies are not routinely performed for diagnosis of EoE owing to its low sensitivity,[2, 44]  but fluoroscopic studies may help in the setting of subtle findings and for evaluation of fibrotic remodeling changes.[43]


Gross Findings

Upper endoscopy for mucosal abnormalities is an integral part of the diagnostic workup in suspected eosinophilic esophagitis (EoE). The endoscopic mucosal changes may be seen through the entire length of the esophagus. The mucosal abnormalities are well described, but none of the features is pathognomonic for EoE and can be seen in other pathological conditions.[2]

The presence of a normal esophageal mucosa on endoscopy does not rule out EoE.[45]

A combination of mucosal abnormalities may be seen, and features vary depending on the stage of the disease. In EoE with active inflammation, the endoscopic findings include diminished vascular pattern, mucosal linear furrows, thick mucosa, and surface exudates. In chronic disease with tissue remodeling, the major endoscopic features seen are fixed rings (also referred to as concentric rings, corrugated esophagus, corrugated rings, ringed esophagus, trachealization), furrows (vertical lines or longitudinal furrows), edema (decreased vascular markings or mucosal pallor), exudates (white spots or plaques), and strictures. Crepe paper esophagus is a minor endoscopic feature due to mucosal fragility.[2, 37, 46, 47]



Microscopic Findings

The normal histology of esophageal mucosa shows a nonkeratinizing stratified squamous epithelium, lamina propria, and muscularis mucosae. The basal cell layer is 1-3 cell layers thick and occupies about 10%-15% of the epithelium. The vascular papillae, which are extensions of lamina propria, extend less than two thirds of the distance from base to surface. Among inflammatory cells, intraepithelial lymphocytes are a normal component of esophageal squamous mucosa. However, eosinophils are not normally seen in esophageal squamous mucosa.

Eosinophilic esophagitis. Normal esophagus. Eosinophilic esophagitis. Normal esophagus.

Histopathologic evaluation plays an integral role in diagnosing eosinophilic esophagitis (EoE). A suspected case of EoE requires histologic confirmation on mucosal biopsy samples. On the same note, the presence of esophageal eosinophilia is not a pathognomonic histological feature of EoE. The presence of eosinophils in esophageal squamous mucosa is most commonly seen in GERD and EoE. Esophageal eosinophilia can also be seen various other pathological conditions, such as achalasia, connective-tissue diseases, vasculitis, drug reactions, and inflammatory bowel disease.[48]

In 2006, a multidisciplinary group at First International Gastrointestinal Eosinophil Research Symposium (FIGERS) proposed a histologic criterion of at least 15 eosinophils/hpf for a diagnosis of EoE based on extensive literature review; this criterion, among other recommendations, was published in 2007.[1, 3]

Eosinophilic esophagitis. Increased eosinophils in Eosinophilic esophagitis. Increased eosinophils in squamous mucosa, more than 15 per high power field (hpf).

Subsequently, the updated recommendations for diagnosis of EoE, published in 2011, stated no change in the threshold number of 15 eosinophils/hpf.[2] This was based on the observation that, since the 2007 consensus recommendations, no studies have identified a clear ‘‘lower limit of esophageal eosinophilia’’ or threshold number that would define EoE or have identified other histologic features or pattern of disease distribution that are pathognomonic of EoE.

Importantly, no change in the use of hpf as the unit of measurement for eosinophilia was made in the 2011 guidelines since no studies have yet determined a standardized size of an hpf. This technical issue is critical, because the size of an hpf can alter the reported number of eosinophils per hpf. The size of a hpf, measured in terms of area, varies from less than 0.1 mm2 to greater than 0.4 mm2, depending on the eyepiece used.[48]

Other pathologic features that support the diagnosis of EoE include degranulated eosinophils, eosinophilic microabscesses (defined as clusters of ≥4 eosinophils), presence of eosinophils in superficial layers of squamous mucosa with or without mucosal sloughing, patchy or diffuse distribution of eosinophils through the entire length of esophageal squamous mucosa, and lamina propria fibrosis.[2, 48, 49] Sloughed off squamous epithelial cells mixed with eosinophils are histologic correlates of white exudates observed on endoscopy.[50]

Eosinophilic esophagitis. Superficial layering of Eosinophilic esophagitis. Superficial layering of eosinophils with desquamation.
Eosinophilic esophagitis. Eosinophilic microabsces Eosinophilic esophagitis. Eosinophilic microabscesses.

EoE is a progressive disease with an inflammatory and fibrostenotic presentation. In the presence of ongoing eosinophilic inflammation, there is esophageal remodeling, leading to subepithelial fibrosis causing esophageal strictures. A histopathologic study by Wang et al that assessed for adequate lamina propria for subepithelial fibrosis in pediatric EoE patients reported that lamina propria fibrosis is patchy and more likely to be detected in middle and distal esophageal biopsies.[51] In this study, most newly diagnosed cases of EoE, had subepithelial fibrosis in esophageal biopsies, despite presenting with inflammatory endoscopic features

The updated consensus guidelines also recommended relaxation of the threshold criteria of eosinophils (>15 eosinophils/hpf) for diagnosis of EoE under certain circumstances, such as when there is a strong clinical evidence of EoE and the biopsy samples show above-mentioned supportive histologic features.[2]

Multiple biopsies from the proximal and distal esophagus are recommended for optimal pathologic evaluation.[2] If biopsy samples from two or more sites show squamous mucosal eosinophil counts of more than 15 per hpf, then a diagnosis “compatible with EoE” may be rendered by the pathologist.[48]

Features of basal cell hyperplasia, dilated intercellular spaces, and elongation of vascular papillae are markers of esophagitis and are not very helpful in distinguishing reflux esophagitis from EoE. However, moderate to marked basal cell hyperplasia occupying more than 50% of mucosal thickness and papillary elongation to 75% of epithelial thickness has been associated with increased eosinophils[52, 53] and with eosinophil degranulation.[54] Esophageal subepithelial fibrosis has been observed to be related to the extent of esophageal eosinophil activation, as evidenced by eosinophil degranulation.[55]

Eosinophilic esophagitis. Eosinophil degranulation Eosinophilic esophagitis. Eosinophil degranulation.
Eosinophilic esophagitis. Architectural abnormalit Eosinophilic esophagitis. Architectural abnormalities (basal cell hyperplasia and elongation of vascular papillae).
Eosinophilic esophagitis. Reflux esophagitis. Eosinophilic esophagitis. Reflux esophagitis.

The presence of erosion, ulceration, and neutrophils are not features of EoE and favor an alternate cause such as reflux esophagitis, infections, or drug-related mucosal damage.



Immunohistochemistry does not play a major role in the diagnosis of eosinophilic esophagitis (EoE), since the predominant inflammatory cell type is an eosinophil, which is easily recognized on hematoxylin and eosin–stained sections. Hence, the recommended diagnostic histological criterion of at least 15 eosinophils/hpf is based on hematoxylin and eosin–stained sections.

However, the use of MBP immunostaining (using monoclonal antibody against MBP) for detection of eosinophils has been shown to increase the yield of detection of eosinophils twofold, and the MBP expression in extracellular areas is thought to reflect foci of eosinophil degranulation.[54]

Immunohistochemical studies have shown an increase in CD3 and CD8 T lymphocytes in the esophageal mucosa of patients with EoE, together with an increase in the antigen presenting CD1a+ Langerhans cell population. Mast cells are also increased.[56, 57]



The etiopathogenesis of eosinophilic esophagitis (EoE) is an evolving concept. Genome expression analysis of patients with EoE has shown a unique transcriptome, which includes several dysregulated genes leading to the inflammatory response.[58] Eotaxin 3 causes tissue eosinophilia. Desmoglein 1 is an intracellular adhesion molecule, which is downregulated in EoE. Interleukin 13 upregulation leads to increased expression of periostin in esophageal fibroblasts. Thymic stromal lymphopoeitin increases basophil responses in EoE.[58]


Prognosis and Predictive Factors

Eosinophilic esophagitis (EoE) is a chronic, antigen-driven inflammatory disease with an allergenic nature of antigen. Current therapies include dietary therapy, topical steroids, and proton pump inhibitors,[59, 60, 61]  as well as endoscopic dilation in cases with strictures.[61, 62] The disease has been seen to remit with treatment such as dietary exclusions of antigenic foods and/or topical steroids.[3, 63, 64]

More recently, investigational therapies include "novel, esophageal-targeted formulas of topic corticosteroids, and monoclonal antibodies (including mepolizumab, reslizumab, QAX576, RPC4046, dupilumab, omalizumab, infliximab and vedolizumab)."[61]

Since EoE is a recently recognized entity, the natural history of EoE and predictive factors of complications are yet unknown.[2]



The main differential diagnosis of eosinophilic esophagitis (EoE) is reflux esophagitis. Other differential diagnoses include primary eosinophilic gastroenteritis.

Reflux esophagitis

This is histological manifestation of GERD that causes esophageal mucosal damage by passive backflow of gastric contents into the distal esophagus. There is considerable histological overlap between reflux esophagitis and EoE. Typically, histological features of reflux esophagitis include mild to moderate basal cell hyperplasia, elongation of vascular papillae, dilated intercellular spaces, increased intraepithelial lymphocytosis, and the presence of intraepithelial eosinophils that number less than 5-7/hpf.

However, in some cases, distal esophageal biopsy samples show a spectrum of overlapping features of reflux esophagitis and EoE. These include moderate to marked basal cell hyperplasia (>50% of mucosal thickness), papillary elongation to 75% of mucosal thickness, dilated intercellular spaces, and the presence of peak intraepithelial eosinophilic counts ranging from 7-14/hpf. In such cases, histological evaluation of additional biopsy samples from the proximal esophagus and midesophagus may be helpful in reaching a more definitive diagnosis.[48, 49]

Primary eosinophilic gastroenteritis

Primary eosinophilic gastroenteritis also shows esophageal eosinophilia along with eosinophilic infiltration in other areas of the gastrointestinal tract . Additional biopsy samples from stomach, duodenum, and colon can help differentiate EoE from primary eosinophilic gastroenteritis.[49]

The main differential diagnosis is reflux esophagitis. Other differential diagnosis includes primary eosinophilic gastroenteritis.


Questions & Answers