Eosinophilic oesophagitis
How eosinophilic inflammation causes dysphagia and food impaction, how biopsy finds EoE, and how PPIs, topical steroids, diet and dupilumab treat it.
- Eosinophilic oesophagitis (EoE) is a chronic, immune-mediated disease of the oesophagus in which a T-helper 2 (Th2) response to food (and sometimes aero-)antigens floods the oesophageal lining with eosinophils.
- It has become a leading cause of solid-food dysphagia and food impaction in both children and adults; prevalence worldwide is now of the order of 40 per 100,000 people. [3]
- Adults usually describe food "going down slowly" or sticking; children present with vomiting, feeding refusal and failure to thrive. The diagnosis rests on endoscopy with biopsies showing at least 15 eosinophils per high-power field. [1][2]
- PPIs, swallowed topical corticosteroids (such as budesonide or fluticasone), and diet are the three established first-line treatments; the biologic dupilumab is reserved for severe, fibrostenotic or treatment-refractory disease. [1][2][3]
- Untreated EoE progresses from inflammation to fixed rings and strictures, so early diagnosis and continued anti-inflammatory treatment matter as much as symptom relief. [4]
By the end of this page you should be able to:
- Explain why type 2 inflammation and barrier dysfunction in EoE produce dysphagia and food impaction.
- Recognise the age-dependent presentation and describe the adaptive eating behaviours adults develop.
- State the diagnostic standard (symptoms plus biopsy) and the biopsy, threshold and endoscopic scoring rules.
- Compare PPI, topical corticosteroid, elimination-diet and dupilumab therapy by suitability, expected histologic response and safety.
- Outline follow-up and endoscopic thresholds for assessing treatment response.
- Adapt management to children, older adults and low-resource settings.
DefinitionEosinophilic oesophagitis (EoE)
A chronic, immune-mediated oesophageal disease characterised by symptoms of oesophageal dysfunction and eosinophil-predominant mucosal inflammation, with at least 15 eosinophils per high-power field on biopsy, and with other causes of oesophageal eosinophilia excluded. PPI (proton pump inhibitor) trials are no longer required to make the diagnosis; PPIs are now a treatment. [1][2]
EoE is not simply reflux. Gastro-oesophageal reflux disease (GORD) can produce mild eosinophilia, but the two conditions coexist and share pathways, and current concept papers treat PPI-responsive oesophageal eosinophilia as a subtype of EoE rather than a separate diagnosis. [4] The driver in EoE is a delayed, non-immunoglobulin E (non-IgE) allergic reaction to antigens, most often foods, with type 2 inflammation and oesophageal remodelling. [2]
EoE has risen from obscurity to a common diagnosis in about three decades. A 2023 systematic review and meta-analysis of population studies reported a pooled global prevalence of roughly 40 per 100,000 people and an incidence of roughly 5.31 per 100,000 person-years. [3] In the general population, histopathology-focused reviews put prevalence at closer to 30 to 52 per 100,000. [4] An Italian pediatric guideline gives Western estimates of up to 20 per 100,000 per year incidence and 1 per 1,000 prevalence, and notes that the annual burden keeps increasing. [1] These figures vary between studies and between adults and children, and the incidence appears to be outpacing improvements in disease recognition. [1][3]
Two demographic patterns matter at the bedside:
- Male sex. Males are affected roughly 2 to 3 times more often than females in pediatric series, and between 2 and 10 times more often in some adult referral populations. [1][4] [2]
- Atopy. Two-thirds of children with EoE have a history of atopy, and EoE sits in the type 2 family with asthma, allergic rhinitis, atopic dermatitis and chronic rhinosinusitis with nasal polyps. [1][2] The odds of these comorbidities in EoE versus controls range from 2.8 to 5.1 in pooled data. [1]
Presentation is also highly conditional on why the endoscopy was done: when endoscopy is performed for dysphagia or food impaction, EoE is found in 63 to 88% of children and in 10 to 15% of adults. [4] That is the clinical clue for a student: in a young person with these symptoms, EoE is not rare.
EoE is multifactorial: genetic susceptibility plus early-life and environmental exposure produces a chronic antigen-driven process.
- Genetic. Risk is familial and twin studies support strong clustering; variants in barrier genes (filaggrin, desmoglein 1, calpain 14) and Th2 signalling molecules (eotaxin-3, periostin, TSLP) are implicated. [3]
- Environmental early-life factors. Formula feeding, cesarean delivery, preterm birth, neonatal intensive care, and early exposure to antibiotics, acid suppressants and histamine-2 receptor blockers have all been associated with later EoE, plausibly through microbiome effects. [1][3]
- Atopic constitution. IgE-mediated food allergy, asthma and atopic dermatitis frequently coexist. [2]
- Antigen. The disease is driven by food antigen exposure; cow's milk, wheat and egg are the most common triggers in dietary studies. [3]
There is no known safe "lifestyle" prevention, and the rising incidence suggests that awareness alone does not explain the whole trend. [1][3]
Everything in EoE follows from three linked steps: a leaky barrier, type 2 inflammation, and progressive remodelling.
- Barrier dysfunction. The oesophageal epithelium in EoE shows reduced desmoglein-1 and E-cadherin, so allergens penetrate more easily; epithelial cells then release alarmins such as thymic stromal lymphopoietin (TSLP), interleukin (IL)-33 and IL-25. [3]
- Th2 inflammation. These alarmins activate antigen-presenting cells and type 2 innate lymphoid cells; naive T cells become Th2 cells that release IL-4, IL-5 and IL-13. IL-13 drives eotaxin-3 (CCL26), which recruits eosinophils via the CCR3 receptor; IL-5 supports eosinophil growth and activation. Mast cells and basophils add histamine and leukotrienes. [3]
- Remodelling. Chronic inflammation leads to basal zone hyperplasia, dilated intercellular spaces, subepithelial fibrosis and eventually fixed rings and strictures; eosinophils can persist focally even when symptoms are mild. [3][4]
This disconnect between symptoms and histology is clinically important. Patients adapt their eating behaviour for years (extra chewing, drinking water with bites, avoiding fibrous foods), so they may not report dysphagia until a stricture is significant. [2] Functional luminal imaging and endoscopic scoring therefore assess disease more objectively than symptom scores alone. [3]
Adults and adolescents usually report:
- solid-food dysphagia, often episodic;
- food impaction, especially with meat and bread;
- adaptive eating behaviour (drinking with meals, cutting food small, prolonged chewing, avoiding hard textures, difficulty swallowing tablets); [2]
- chest pain or heartburn.
Infants and young children present more vaguely with reflux-like symptoms, vomiting, nausea, abdominal pain, food refusal and failure to thrive; dysphagia and impaction appear in school-age children. [1]
Atopic background. Asthma, allergic rhinitis, atopic dermatitis or IgE-mediated food allergy is common; a family history of dysphagia or EoE also raises suspicion. [2]
There is often nothing to find physically. The value is in the dietary and atopic history and in testing for coexisting allergic disease. Weight may be low in children. Abdominal and chest examinations are usually normal unless a complication has occurred.
A food bolus stuck in the oesophagus, or the inability to swallow saliva, is an emergency needing endoscopic removal; taking oesophageal biopsies at the same time is strongly encouraged. [2]
| Condition | What separates it from EoE |
|---|---|
| Gastro-oesophageal reflux disease | Milder, more diffuse eosinophilia; responds to acid suppression; endoscopic EoE features may be absent. Reflux and EoE can coexist, and PPI-responsive eosinophilia is now classed within EoE. [4] |
| Eosinophilic gastrointestinal disease (non-oesophageal) | Eosinophilia extends into stomach or duodenum; biopsy those sites at the first endoscopy. [1] |
| Achalasia and other motility disorders | Absent primary peristalsis, dilated oesophagus; biopsy threshold not reached. [2] |
| Drug hypersensitivity | Temporal link to a drug; may need withdrawal and re-biopsy. [4] |
| Crohn disease affecting the oesophagus | Granulomas, transmural disease, disease elsewhere. [1] |
| Hypereosinophilic syndrome | Massive peripheral eosinophilia, multi-organ disease. [1] |
| Structural disease (web, ring, tumour) | Discrete lesion on endoscopy or imaging; biopsy shows no eosinophilic infiltrate. |
Endoscopy is required for diagnosis regardless of how normal the mucosa looks. Characteristic but not mandatory findings include oedema, fixed rings, white exudates, linear furrows, strictures and fragile "crepe-paper" mucosa. [2][4] The standardised way to report them is the EoE Endoscopic Reference Score (EREFS), which splits features into inflammatory (oedema, exudates, furrows) and fibrostenotic (rings, strictures) components. [2][3][5]
When EREFS is used to monitor treatment, improvement can be quantified: in the LIBERTY EoE TREET programme, dupilumab 300 mg weekly improved the worst-observed-region EREFS total score by a mean of 2.17 points compared with placebo at 24 weeks. [5]
- Threshold: at least 15 eosinophils per high-power field (per 0.3 mm2; equivalent to 60 eosinophils per mm2 in the international consensus definition). [1][2][4]
- Sampling: a minimum of 6 biopsies from at least 2 levels (distal, mid and proximal oesophagus); Canadian consensus phrasing asks for at least 2 biopsies per level from at least 2 segments, giving 4 to 6 in total. [1][2]
- Ancillary findings that strengthen the diagnosis: eosinophil microabscesses (clusters of at least 4 eosinophils), superficial layering, basal zone hyperplasia, dilated intercellular spaces, dyskeratotic cells and lamina propria fibrosis. [4]
- Scoring: the Histologic Scoring System (HSS) grades eight features from 0 (normal) to 3 (maximum), to a maximum of 24, capturing inflammation and remodelling more completely than eosinophil counts alone. [4]
- PPI handling: stop PPIs 3 to 4 weeks before biopsy collection to avoid transient suppression of eosinophilia. [1]
- Allergy testing (skin prick, specific IgE, patch tests) is useful for concurrent IgE-mediated food allergy and atopic comorbidity, but should not be used to select empirical elimination diets for EoE: it has low to moderate predictive value for true EoE trigger foods. [1][2][3]
- Blood tests and imaging are not diagnostic but help exclude mimics and assess nutrition.
- Non-invasive tools such as the oesophageal string test, Cytosponge and transnasal endoscopy are emerging and are reserved to centres with experience; they do not yet replace biopsy. [3]
- Motility testing is reserved for atypical presentations or poor response.
The diagnosis requires all three elements: symptoms of oesophageal dysfunction, oesophageal biopsy with at least 15 eosinophils per high-power field, and exclusion of other causes of eosinophilia. [1][2][4] A PPI trial is not needed for diagnosis. [1][2]
For historical context, high-dose PPI taken for 8 to 12 weeks can still be used as a therapeutic trial (remission in 30 to 50% of patients), because PPI-responsive oesophageal eosinophilia is now considered part of the EoE spectrum and PPIs have anti-inflammatory effects beyond acid suppression. [3][4]
The three elements of care are: control symptoms, suppress inflammation, and watch for fibrostenotic change. Shared decision-making with the patient (and, for children, the family) is central, because drug, diet and procedural options differ in efficacy, burden and acceptability.
Approach to a patient with suspected EoE
- Suspect EoESolid-food dysphagia, food impaction, or vomiting and feeding refusal in a child; often with atopy
- Emergency firstFood bolus obstruction or inability to swallow saliva: urgent endoscopic removal, with oesophageal biopsies at the same time
- Upper GI endoscopy with at least 6 biopsiesFrom distal, mid and proximal oesophagus; include stomach and duodenum if non-oesophageal symptoms; stop PPI 3-4 weeks beforehand
- Question: Biopsy shows at least 15 eosinophils per high-power field and other causes excluded?
No
- Reconsider mimicsGORD, motility disease, structural lesion, drug hypersensitivity; repeat biopsy if PPI effect suspected
Yes: EoE confirmed
- Stage phenotypeInflammatory vs fibrostenotic; EREFS and HSS scores; assess strictures and oesophageal calibre
- Choose first treatmentMild disease: PPI; inflammatory phenotype: swallowed topical steroid; atopic child or trigger likely: elimination diet starting with cow's milk; severe fibrostenosis or first-line failure: dupilumab
- Endoscopy at 8-12 weeks (ideally 6-12)Check histologic remission (6 or fewer eosinophils per high-power field), EREFS and symptoms; single therapy first where possible
- Long-termContinue effective therapy; monitor every 1-2 years even if asymptomatic; dilate strictures as needed
| Option | Best suited to | Reported histologic remission | Key cautions |
|---|---|---|---|
| Proton pump inhibitor | Mild disease; patients with reflux symptoms; young children | 30 to 50% within 8 to 12 weeks. [3] | Weight-based paediatric dosing is 2 mg/kg/day; adult dosing follows local high-dose regimens. Acid-independent anti-inflammatory effect is shared across the class. [1] |
| Swallowed topical corticosteroid | Inflammatory phenotype; adults needing rapid mucosal control | 60 to 95% in children and adults. [3] | Budesonide orodispersible tablet is licensed in Canada for adults; viscous budesonide and inhaled fluticasone swallowed from an MDI are off-label. [2] Systemic or long-term oral corticosteroids are not recommended. [2] |
| Elimination diet | Atopic children, patients preferring non-drug therapy | About 51% with one-food, 49% with four-food, 61% with six-food elimination; elemental diets 90 to 95%. Overall 54%. [3] | Start with one food (cow's milk) or two (milk plus wheat); counsel with a dietitian. Six-food removal is not first-line despite higher remission, because of burden and adherence. [2][3] Allergy testing should not guide empiric elimination. [2][3] |
| Dupilumab | Severe, fibrostenotic or treatment-refractory EoE, or concurrent type 2 disease | Histologic remission in about 60% versus 5% on placebo in LIBERTY EoE TREET; endoscopic improvements sustained to 52 weeks. [3][5] | Licensed from age 1 year and body weight at least 15 kg, with weight-based dosing (200 mg every 2 weeks for 15 to under 30 kg; 300 mg every 2 weeks for 30 to under 40 kg; 300 mg once weekly for 40 kg or more); usually reserved for failure of PPI, topical steroid or diet. [6] |
PPIs are the preferred first-line in mild disease in the Italian 2025 pediatric guideline, at 2 mg/kg/day. [1] They work partly by reducing acid, and partly through anti-inflammatory effects on the oesophageal epithelium; symptom and histologic remission rates are 30 to 50% within 8 to 12 weeks. [3] In adults, high-dose regimens (often twice daily) are used; exact drug and dose should follow local guidance rather than a single global rule. Stopping PPIs is a shared decision because maintenance helps prevent fibrosis. [2][4]
These deliver drug to the mucosa with minimal systemic absorption and are the most reliably effective first-line class for histologic remission (60 to 95%). [3] Options include:
- budesonide orodispersible tablet, approved in Canada for adults; [2]
- viscous swallowed budesonide (off-label); [2]
- fluticasone metered-dose inhaler with the dose swallowed (off-label). [2]
Oral systemic corticosteroids are not recommended for routine EoE, because relapse is rapid and steroid toxicity follows long-term use. [2]
Because EoE is antigen-driven, removing the offending food can induce remission. Current consensus recommends a step-up approach rather than starting with six-food removal: begin with one (cow's milk) or two foods (milk plus wheat), then step up if needed, with dietitian support. [2][3] In the pooled data, cow's milk was the most common trigger (70.5%), followed by wheat (48.0%) and egg (27.6%); overall histologic remission with empirical elimination was 53.8%. [3] Elemental amino-acid formula achieves 90 to 95% remission but is unpalatable, expensive and usually requires tube feeding, so it is reserved for refractory cases. [3]
Prolonged dietary exclusion carries a real hazard in atopic patients: avoiding a food for long periods can permit IgE-mediated sensitisation and anaphylaxis on re-exposure, so transition periods and re-introduction should be supervised and allergy input sought where needed. [2]
Dilation relieves a stricture or narrow calibre but does not treat the underlying inflammation; it should therefore be used as an adjunct only, with medical or dietary therapy continued. [2] Most narrowings can be dilated, but perforation risk, although small, is real in a fragile oesophagus.
Dupilumab blocks the IL-4/IL-13 axis through the IL-4 receptor alpha subunit. It is licensed for EoE from 1 year of age in patients weighing at least 15 kg, at weight-based doses (200 mg every 2 weeks below 30 kg; 300 mg every 2 weeks for 30 to below 40 kg; 300 mg once weekly for 40 kg or more), and is generally reserved for disease refractory to, or intolerant of, standard therapy, or where another approved severe type 2 condition is present. [6] In LIBERTY EoE TREET it produced histologic remission in 60% versus 5% on placebo, with large peak eosinophil reductions and sustained dysphagia, EREFS and dispensability improvements through 52 weeks; post hoc data show improvements in both inflammatory and fibrostenotic endoscopic components. [3][5]
Endoscopy with biopsies should be repeated after a change in management, ideally at 6 to 12 weeks (many guidelines phrase 8 to 12 weeks). [1][2] Maintenance treatment is typically continued for years; the optimal duration once remission is achieved is not well defined and is guided by severity, phenotype and shared decision-making, because fibrostenotic risk accumulates silently. [2][4] Long-term monitoring every 1 to 2 years is advised even when symptoms have resolved. [1]
- Strictures and a narrow-calibre oesophagus from fibrostenotic remodelling; risk rises with untreated or delayed diagnosis. [2][4]
- Food impaction, sometimes needing endoscopic removal, and rarely perforation. [2][4]
- Growth failure and feeding aversion in children. [1]
- Malnutrition and vitamin or micronutrient deficiency on very restrictive diets. [2]
- Psychosocial burden of chronic dysphagia, including anxiety around eating. [2]
With appropriate therapy EoE usually achieves symptom and histologic control, and a disease-modifying strategy can slow or stop fibrostenotic progression, although long-term remission varies between individuals. [2][3] Cessation of treatment usually leads to recurrent inflammation and remodelling, which is why EoE is managed as a chronic disease rather than treated episodically. [3]
- Children. School presentation may be non-specific. Endoscopy should be performed in a pediatric setting where possible. Dietary steps dominate in atopic children; weight-based PPI dosing (2 mg/kg/day) applies. [1]
- Adolescents transitioning to adult care. This is a high-risk period for loss to follow-up and worsening fibrosis. [2]
- Older adults. Fibrostenotic phenotypes and comorbid reflux are more common, and polypharmacy changes PPI interactions; endoscopic dilation may be needed earlier. [2][4]
- Pregnancy. Safety data for swallowed topical steroids and PPIs are reassuring, and dupilumab's use in pregnancy should be discussed with obstetrics and gastroenterology; the priority is maintaining nutrition and avoiding impaction. [2]
East Africa
Endoscopy and histopathology are not available everywhere, so early recognition depends on the history: solid-food dysphagia in an atopic young person, and recurrent food impaction. Where a flexible endoscope exists, take 6 biopsies from different levels even if the mucosa looks normal, and process them with the standard eosinophil count. PPIs, viscous steroid suspensions and step-up diets are relatively inexpensive, whereas dupilumab, budesonide orodispersible tablets and enteral formula diets are often out of reach; dilation of strictures and dietary counselling may then carry most of the benefit. Referral to a centre with EoE expertise, and a low threshold for nutritional support, improve outcomes.
- EoE is a chronic type 2 immune-mediated oesophageal disease driven by barrier dysfunction and allergic (usually food) antigen exposure.
- Adults present with solid-food dysphagia, food impaction and adaptive eating behaviour; children present with vomiting, feeding refusal and failure to thrive.
- Prevalence has risen to about 30 to 52 per 100,000, and the disease is roughly 2 to 3 times more common in males.
- Diagnosis requires symptoms plus at least 15 eosinophils per high-power field on biopsy; at least 6 biopsies from 2 or more levels are taken even if the mucosa looks normal.
- A PPI trial is no longer needed for diagnosis; PPIs (30 to 50% histologic remission at 8 to 12 weeks) are a first-line treatment.
- Swallowed topical corticosteroids achieve 60 to 95% histologic remission; oral systemic steroids are avoided.
- Diet is antigen-directed; step up from one food (cow's milk) or two, with dietitian support; elemental diet is reserved for refractory disease.
- Dupilumab is licensed for EoE from 1 year of age (body weight at least 15 kg), with weight-based dosing (300 mg weekly only at 40 kg or more), and is generally reserved for refractory or fibrostenotic cases. [6]
- Endoscopy is repeated 6 to 12 weeks after a change of treatment; monitoring continues every 1 to 2 years even if asymptomatic.
- Untreated EoE progresses from inflammation to fixed rings, strictures and food impaction, so treat early and continue therapy.
Answer each one in your head first, then open the answer.
A 24-year-old man reports 'food sticking' for 3 years and has learned to chew for a long time and drink water with meals. He has allergic rhinitis. Upper endoscopy is normal-looking. What is the most likely diagnosis and what must be done during endoscopy?
Show answerHide answer
Eosinophilic oesophagitis. Even with a normal-appearing mucosa, at least 6 oesophageal biopsies from 2 or more levels (distal, mid, proximal) must be taken, with the patient off PPIs for 3 to 4 weeks, and should show at least 15 eosinophils per high-power field to confirm it. EREFS should also be scored. [1][2][4]
Why might a patient with significant histologic EoE have few symptoms?
Compare the typical first-line treatment of inflammatory versus fibrostenotic EoE.
Show answerHide answer
Inflammatory disease usually responds well to swallowed topical corticosteroids (60 to 95% histologic remission) or PPIs (30 to 50% at 8 to 12 weeks); diet is useful in atopic patients. Once fixed rings or strictures dominate, dilation helps the lumen but medical or dietary therapy must continue to prevent recurrence, and dupilumab (licensed from 1 year of age, at least 15 kg, at weight-based doses) is considered for refractory or fibrostenotic disease. [6]
A 7-year-old atopic child with vomiting and failure to thrive is diagnosed with EoE. Why might an elimination diet be preferred and why should it start with one or two foods?
Show answerHide answer
Diet addresses the antigen driving disease in atopic children. Current evidence favours step-up empiric removal rather than 6-food restriction: begin with cow's milk or milk plus wheat, because it balances efficacy (about 51% and 49% remission respectively) with adherence and quality of life, whereas 6-food removal is more burdensome; dietitian support is essential. [2][3]
Is a PPI trial required to diagnose EoE in a child whose dysphagia improves on a PPI?
Show answerHide answer
No. PPI-responsive oesophageal eosinophilia is now considered part of the EoE spectrum, and PPI trials are no longer a diagnostic requirement; PPIs are used as a treatment (2 mg/kg/day in children), with biopsies taken 3 to 4 weeks after stopping them. [1]
Which endoscopic features make up the EREFS score and how are they grouped?
Show answerHide answer
- Gastro-oesophageal reflux disease
- Functional dysphagia
- Atopic disorders and type 2 inflammation
- Endoscopic biopsy and histologic scoring
- Food allergy and elimination diets
- Oliva S, Arrigo S, Bramuzzo M, et al. Eosinophilic esophagitis in children and adolescents: a clinical practice guideline. Italian Journal of Pediatrics 2025;51:242. doi:10.1186/s13052-025-02056-x. Accessed .
- Avinashi V, Gupta M, Payne BA, et al. Recommendations for the diagnosis and management of eosinophilic esophagitis in adults and children in Canada: a Delphi consensus project. Journal of the Canadian Association of Gastroenterology 2025;8:207-227. doi:10.1093/jcag/gwaf022. Accessed .
- Sharma A, Tiwari A, Deshpande V, et al. Advancing the understanding of eosinophilic esophagitis: From pathogenesis to novel therapies. World Journal of Gastrointestinal Pathophysiology 2026;17:121571. doi:10.4291/wjgp.121571. Accessed .
- Salzano S, Caltabiano R, Zizzo M, et al. Eosinophilic Esophagitis: A Review of Histopathological and Diagnostic Aspects. Diagnostics (Basel) 2025;15:3011. doi:10.3390/diagnostics15233011. Accessed .
- Dellon ES, Collins MH, Peterson K, et al. Dupilumab Improves Endoscopic Features in Adult and Adolescent Patients with Eosinophilic Esophagitis: Post Hoc Analysis of LIBERTY EoE TREET Study. Clinical and Translational Gastroenterology 2026;17:e01090. doi:10.14309/ctg.0000000000001090. Accessed .
- Regeneron Pharmaceuticals Inc. and Sanofi. DUPIXENT (dupilumab) injection, for subcutaneous use: prescribing information. U.S. Food and Drug Administration prescribing information, revised 2025. EoE indication: adults and paediatric patients 1 year of age and older weighing at least 15 kg; dosage table.. Accessed .
Educational content for medical students and doctors. It is not a substitute for clinical judgement or local guidelines.