From allergen exposure to histologic diagnosis of eosinophilic esophagitis
Eosinophilic esophagitis (EoE) is a chronic, antigen-driven, immune-mediated disease of the esophagus. Unlike IgE-mediated food allergy, EoE is largely a Th2-type, cell-mediated (delayed) hypersensitivity response to swallowed food proteins and, to a lesser extent, inhaled aeroallergens, occurring in a genetically and barrier-predisposed esophageal mucosa.
The healthy esophageal squamous epithelium is a tightly sealed barrier: desmosomal proteins (desmoglein-1, desmoplakin) and tight-junction complexes keep luminal antigens out of the lamina propria. In EoE, genetic variants (notably near CAPN14, a gene encoding an esophageal-specific calpain) combine with IL-13 signaling to loosen this barrier, allowing dilated intercellular spaces to open between squamous cells.
Through this leaky barrier, common dietary proteins — cow's milk, wheat, egg, soy, peanut/tree nut, and fish/shellfish account for the large majority of triggers — penetrate the epithelium. Aeroallergens (pollens) can also provoke or worsen disease, explaining why some patients show seasonal symptom variation.
Unlike classic IgE-mediated food allergy (urticaria, anaphylaxis within minutes), EoE antigen responses are delayed and chronic, driven predominantly by CD4+ Th2 lymphocytes rather than mast-cell degranulation, although mixed IgE/non-IgE mechanisms are increasingly recognized.
EoE was first formally described as a distinct clinicopathologic entity in the early 1990s; incidence has risen roughly 5- to 10-fold since then, outpacing ascertainment bias alone — a pattern shared with other atopic diseases and attributed in part to the "hygiene hypothesis" and changing microbial exposures.
Esophageal dendritic cells and Langerhans-like cells extend processes between epithelial cells to sample luminal antigen. After antigen capture, dendritic cells migrate to regional lymph nodes and present processed peptide to naive CD4+ T cells via MHC class II.
In a Th2-permissive cytokine milieu — shaped by epithelial-derived alarmins such as thymic stromal lymphopoietin (TSLP), IL-25, and IL-33, all released by the injured epithelium — naive T cells differentiate preferentially into Th2 effector cells rather than Th1 or Th17 lineages.
Activated Th2 cells secrete a defining cytokine triad: • IL-4: drives further Th2 differentiation and B-cell class switching toward IgE • IL-13: the dominant driver of the esophageal EoE transcriptome — induces eotaxin-3, barrier gene suppression, and basal cell hyperplasia • IL-5: the principal survival and maturation factor for eosinophils, acting on bone marrow and circulating eosinophil precursors
Mast cells, basophils, and innate lymphoid cells (ILC2s) amplify this Th2 circuit, and local IgE production against food antigens has been demonstrated directly within esophageal tissue in some patients.
EoE is a polygenic disease. Genome-wide association studies have implicated loci at CAPN14 (esophageal epithelial barrier/differentiation), TSLP and its receptor, and 2p23 near the eotaxin-3 (CCL26) gene itself, among others. Family history increases risk roughly 10-fold, and monozygotic twin concordance exceeds dizygotic concordance, confirming a heritable component layered on environmental triggers.
Early-life factors associated with increased EoE risk mirror those linked to other allergic diseases: cesarean delivery, antibiotic exposure in infancy, and lack of breastfeeding — all consistent with a microbiome-mediated modulation of mucosal immune tolerance.
Once Th2 polarization is established, a cytokine relay converts the esophageal mucosa into a tissue that actively recruits eosinophils from the bloodstream. This step — chemokine-driven transendothelial migration into the epithelium — is the direct cellular event that produces the histologic hallmark of EoE.
IL-5, secreted by Th2 cells and ILC2s, is the master regulator of eosinophil lineage commitment, maturation, and survival. It acts on CD34+ bone-marrow progenitors to accelerate eosinophilopoiesis and prolongs peripheral eosinophil lifespan by suppressing apoptosis.
Circulating eosinophil counts are often — but not always — elevated in EoE (peripheral blood eosinophilia is present in a minority to about half of patients, so a normal blood count does not exclude the disease). The clinically decisive eosinophilia is tissue-based, localized to the esophageal mucosa, not systemic.
This IL-5 axis is directly therapeutic: anti-IL-5 (mepolizumab, reslizumab) and anti-IL-5Rα (benralizumab) biologics, developed originally for eosinophilic asthma, reduce blood and tissue eosinophils but have shown only modest or inconsistent histologic and symptomatic benefit in EoE trials — underscoring that eosinophils are one contributor among several (Th2 cytokines, mast cells, remodeling) to the disease phenotype.
While IL-5 supplies eosinophils systemically, the tissue-specific chemokine eotaxin-3 (CCL26) — produced by IL-13-stimulated esophageal epithelial cells and fibroblasts — is what actually pulls eosinophils out of the vasculature and into the epithelium via the CCR3 receptor expressed on eosinophils.
CCL26 is the single most upregulated gene in EoE esophageal biopsies compared with normal mucosa, forming the backbone of the "EoE transcriptome" (a panel of ~90–300 differentially expressed genes) used in research and in an experimental minimally invasive diagnostic tool, the esophageal string test / cytosponge, that samples luminal contents without formal endoscopy.
Eosinophils cross the vascular endothelium by rolling (selectins), firm adhesion (integrins, VCAM-1), and diapedesis, then migrate along the CCL26 concentration gradient into the papillae of the lamina propria and finally into the squamous epithelium itself — a compartment that is entirely eosinophil-free in healthy individuals.
Eosinophils are not passive bystanders: their granules contain major basic protein (MBP), eosinophil cationic protein (ECP), eosinophil peroxidase, and eosinophil-derived neurotoxin — all directly cytotoxic to epithelial cells and capable of activating fibroblasts, contributing to the smooth-muscle dysfunction and subepithelial fibrosis that drive strictures in long-standing, untreated disease.
IL-13 additionally suppresses desmoglein-1 and filaggrin expression in the epithelium, widening intercellular spaces and further increasing antigen penetration — creating a self-reinforcing cycle: more antigen exposure → more Th2 activation → more IL-13/eotaxin-3 → more barrier breakdown → more antigen exposure.
Over months to years, chronic unresolved inflammation drives basal zone hyperplasia (basal cells expand from the normal <15% of epithelial thickness to over 50%), dilated intercellular spaces visible on light microscopy, and eventually lamina propria fibrosis — the structural substrate for the rings and strictures seen endoscopically in the next stage.
Upper endoscopy is required in every suspected EoE case, both to inspect the mucosa and to obtain biopsies. Endoscopic appearance alone cannot diagnose EoE (biopsy is mandatory), but a standardized scoring system — EREFS — quantifies the inflammatory and fibrostenotic findings, tracks disease severity, and helps target biopsy sites.
EREFS (Edema, Rings, Exudates, Furrows, Stricture) is a validated, reproducible endoscopic reference score used to grade EoE severity at every exam:
• Edema: loss of the normal vascular (branching) pattern and mucosal pallor, reflecting tissue swelling — often the most subtle finding and easy to miss • Rings (trachealization / feline esophagus): concentric circular folds that make the esophagus resemble tracheal cartilage rings, from chronic mural inflammation and fibrosis of circular smooth muscle • Exudates: white plaques or specks scattered on the mucosa — histologically these are eosinophil microabscesses, not infection (do not confuse with Candida) • Furrows: longitudinal linear grooves running the length of the esophagus, from edema of the mucosal folds • Stricture: fixed luminal narrowing from long-standing subepithelial fibrosis — the end-stage fibrostenotic consequence of unresolved inflammation, occurring in a substantial minority of adults, especially with delayed diagnosis
A related but distinct sign, "crepe-paper mucosa," describes friable, easily torn epithelium that lacerates with passage of the endoscope or with dilation — a marker of severe, fragile fibrotic disease.
EREFS severity correlates with histologic eosinophil density and with esophageal distensibility measured by functional luminal imaging probe (FLIP), but a normal-appearing esophagus does NOT rule out EoE — endoscopic findings are absent in an estimated one in four to one in ten biopsy-proven cases, which is precisely why biopsies are mandatory regardless of visual appearance.
Symptoms differ sharply by age. In young children: feeding difficulty, food refusal, vomiting, failure to thrive, and abdominal pain predominate. In older children and adolescents: heartburn-like symptoms and vomiting persist alongside emerging dysphagia. In adults: solid-food dysphagia and esophageal food bolus impaction dominate — EoE is now recognized as the leading cause of food impaction requiring emergency endoscopic removal in younger adults presenting with this event.
Adults with long-standing, undiagnosed EoE often unconsciously adapt — chewing extensively, avoiding dry or fibrous foods, drinking liquids with every bite — masking symptom severity and contributing to the historically long lag between symptom onset and diagnosis.
Esophageal eosinophilia is not exclusive to EoE. Before a diagnosis is finalized, clinicians must exclude other causes of secondary esophageal eosinophilia: gastroesophageal reflux disease, eosinophilic gastrointestinal diseases beyond the esophagus, achalasia and other motility disorders, connective tissue disease, hypereosinophilic syndrome, infection, drug hypersensitivity reactions, and Crohn disease involving the esophagus.
Historically, patients whose esophageal eosinophilia and symptoms resolved with proton pump inhibitor (PPI) therapy were labeled as having a separate condition, "PPI-responsive esophageal eosinophilia" (PPI-REE). Updated international consensus guidelines (AGREE, 2018) reclassified PPI-REE as part of the EoE spectrum rather than a distinct disease, since these patients share the same clinical, endoscopic, histologic, and molecular (transcriptomic) profile as PPI non-responders — PPI is now considered a first-line EoE treatment, not a diagnostic exclusion test.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Edema | Grade 0–1 | Loss of vascular markings / mucosal pallor | Often the earliest, subtlest finding |
| Rings | Grade 0–3 | Mild transient (0–1) to fixed non-passable stenosis (3) | Correlates strongly with fibrosis burden |
| Exudates | Grade 0–2 | None; <10% of surface; >10% of surface | Represents eosinophilic microabscesses |
| Furrows | Grade 0–1 | Absent vs. present longitudinal grooves | Reflects mucosal/submucosal edema |
| Stricture | Grade 0–1 | Absent vs. present luminal narrowing | Marker of chronic, fibrostenotic disease |
Because eosinophilic infiltration in EoE is histologically patchy — dense in one field, sparse or absent a few millimeters away — a single biopsy carries a substantial risk of a false-negative result. Guideline-directed sampling from multiple esophageal levels is what makes biopsy-based diagnosis reliable.
Landmark biopsy-mapping studies established that eosinophil density varies markedly along the length of the esophagus and even within a single 2 mm segment — some fields exceed 50 eosinophils/hpf while adjacent fields show none. Sampling only the distal esophagus, as was common practice when EoE was still confused with reflux esophagitis, misses a clinically significant fraction of cases because reflux disease classically affects only the very distal esophagus while EoE is panesophageal.
Current society guidelines (ACG, AGA/joint task force consensus updates) recommend a minimum of six biopsies distributed across at least two esophageal levels — typically proximal/mid and distal — obtained regardless of whether the intervening mucosa looks endoscopically normal, since normal-appearing mucosa can still harbor diagnostic eosinophil counts.
Biopsy targeting matters as much as biopsy number: forceps should preferentially sample visibly abnormal areas — furrows, exudates, rings — where eosinophil density is highest, while still including at least one bite from each level even if it looks unremarkable, to avoid missing patchy disease in a grossly normal-appearing segment.
Standard flexible upper endoscopy under sedation is used; biopsy forceps obtain superficial pinch samples (typically capturing mucosa and superficial lamina propria, not full-thickness wall). Because chronic EoE tissue can be fragile and friable ("crepe-paper mucosa") from fibrosis, biopsy and any concurrent dilation carry an increased, though still low, risk of mucosal laceration or perforation compared with biopsy of normal esophageal tissue — a risk that rises further if dilation is performed in the same session as active, unresolved inflammation.
Biopsies should also be taken from the gastric antrum and duodenum in the initial workup when eosinophilic gastrointestinal disease beyond the esophagus is a consideration, since EoE is defined as an isolated esophageal process and coexisting eosinophilic gastritis/enteritis changes management.
Histology is read for a constellation of features supportive of EoE, not eosinophil count alone:
• Eosinophil microabscesses: clusters of ≥4 eosinophils, correlating with the endoscopic exudates • Surface layering: eosinophils concentrated toward the luminal surface rather than evenly distributed • Basal zone hyperplasia: basal (proliferative) layer expanded beyond ~15% of total epithelial thickness • Dilated intercellular spaces: a marker of barrier disruption, visible as widened gaps between squamous cells • Lamina propria fibrosis: increased subepithelial collagen deposition, present when biopsies are deep enough to include lamina propria — a marker of remodeling and risk for stricture
These secondary features help distinguish EoE from other causes of esophageal eosinophilia and support disease-activity assessment even when eosinophil counts hover near the diagnostic threshold.
The diagnosis of EoE requires the convergence of three elements: compatible esophageal symptoms, compatible endoscopic and/or histologic findings, and a peak eosinophil count of ≥15 per high-power field on biopsy, after other causes of esophageal eosinophilia have been reasonably excluded. Treatment aims to reduce that count back toward histologic remission.
The pathologist scans all submitted biopsy fragments at low power to identify the area of densest eosinophilic infiltrate, then counts eosinophils in that single most-involved high-power field (400× magnification, roughly 0.3 mm² depending on the microscope's field-of-view number). The single highest count across all fragments and all levels — the "peak eosinophil count" — is what is compared against the diagnostic threshold.
A peak count ≥15 eosinophils/hpf, in the correct clinical and endoscopic context, meets histologic criteria for EoE. Counts below 15/hpf do not exclude EoE if biopsies were poorly targeted, too few in number, or taken during a period of low disease activity (e.g., after inadvertent dietary avoidance) — which is exactly why the multilevel protocol from the previous stage matters so much for a reliable negative result.
The ≥15 eos/hpf cutoff was empirically derived and validated against reflux esophagitis and other secondary eosinophilic conditions, which rarely exceed single-digit counts confined to the very distal esophagus.
Symptom severity correlates poorly with eosinophil count, especially in adults, who may develop esophageal remodeling and fibrosis that causes dysphagia even when active inflammation (and eosinophil count) has substantially improved — this is why guidelines now emphasize repeat endoscopic biopsy, not symptoms alone, to confirm histologic response to treatment.
Three foundational treatment strategies, and one newer biologic class, are used alone or in combination, guided by shared decision-making:
• PPI trial: omeprazole or equivalent, typically twice daily for 8 weeks, now considered a first-line EoE therapy (not merely a diagnostic exclusion step) — induces histologic remission in a substantial proportion of patients, plausibly via anti-inflammatory effects on epithelial gene expression independent of acid suppression • Topical swallowed corticosteroids: fluticasone propionate (metered-dose inhaler sprayed into the mouth and swallowed) or budesonide (oral viscous slurry or effervescent tablet), typically for 8–12 weeks — achieves histologic remission in roughly half to three-quarters of patients in trials • Dietary elimination: empiric removal of the most common trigger foods. The historical six-food elimination diet (milk, wheat, egg, soy, nuts, fish/shellfish) achieves remission in a majority of patients but is burdensome; a step-up approach starting with one- or two-food elimination (typically dairy ± wheat) is now favored, escalating only if needed, followed by systematic food reintroduction with repeat endoscopy to identify the specific trigger(s) • Dupilumab: a monoclonal antibody against IL-4Rα, the shared receptor subunit for IL-4 and IL-13, blocking both cytokines simultaneously — FDA approved for EoE in 2022 for patients aged 12 and older weighing ≥40 kg, with the indication later extended to younger children; pivotal trials showed histologic remission in roughly 60% of treated patients versus a low single-digit percentage on placebo
Because symptoms and histology can diverge, guidelines recommend repeat upper endoscopy with biopsy roughly 8–12 weeks after starting or changing therapy to objectively confirm histologic response, rather than titrating treatment on symptoms alone. The same multilevel biopsy protocol is repeated at follow-up.
Untreated or under-treated chronic EoE risks progressive esophageal remodeling: lamina propria fibrosis, reduced esophageal distensibility (measurable by FLIP), stricture formation, and recurrent food impactions requiring endoscopic dilation. Esophageal dilation itself does not treat the underlying inflammation and is reserved as an adjunct for fibrostenotic complications once anti-inflammatory therapy has been optimized.
EoE is a chronic disease: discontinuing effective therapy typically leads to histologic and symptomatic relapse, so long-term maintenance treatment — at the lowest effective dose or with continued dietary avoidance of an identified trigger — is generally required indefinitely.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| PPI trial | All age groups, first-line | Acid suppression + direct anti-inflammatory gene effects; 8-week trial | Oral, low cost, no diet restriction |
| Topical corticosteroids | Fluticasone / budesonide | Local mucosal anti-inflammatory action, minimal systemic absorption | High remission rates, non-systemic |
| Dietary elimination | 1-, 2-, 4-, or 6-food elimination | Removes causative food antigen(s) identified by reintroduction | Addresses root cause, drug-free |
| Dupilumab (biologic) | Moderate–severe, refractory disease | Anti-IL-4Rα blocks IL-4 and IL-13 signaling | FDA-approved, strong trial efficacy |