Subscribe
Share
Share
Embed
Every diagnostic criterion and every drug in the EoE ladder is downstream of one mechanism: a Th2 allergic response to food allergen in a susceptible host that ends in fibrosis, rings, and impaction if untreated. Hold the cytokine story and the 15-eosinophil threshold, and the diagnosis, the EREFS score, and the treatment ladder follow.
The case. A 32-year-old atopic man presents with a food bolus impaction. After it is cleared, what biopsies do you take, how many, and what threshold makes the diagnosis?
Topics covered
Key decisions
For the full chapter with MCQs, tables, and primary-guideline references, visit www.boardpearls.com.
Questions or feedback: hello@boardpearls.com.
Board Pearls is a gastroenterology board review built around clinical reasoning, not recall. Each episode takes one high-yield topic and works it the way you would on rounds: a case to anchor it, the framework that sorts the differential, and the specific decisions the exam rewards.
The gastroenterology series covers the full blueprint across nine modules: esophagus, stomach and duodenum, small bowel, colon, pelvic floor, liver, pancreas and biliary, endoscopy, and the cross-cutting topics. Episodes are grouped by chapter and built from the primary guidelines and pivotal trials the boards draw from (ACG, AGA, AASLD, ASGE), not from textbook summaries.
Use it as an audio companion to the written curriculum, MCQs, and AI tutor at boardpearls.com. Questions or feedback: hello@boardpearls.com.
Welcome to Board Pearls. This is episode one of two of the Eosinophilic and Infectious Esophagitis chapter, in the Esophageal Disorders module. In this episode we cover eosinophilic esophagitis from diagnosis through refractory disease: a Th2 allergic disease defined by fifteen eosinophils per high-power field on esophageal biopsy, the EREFS endoscopic score, the treatment ladder of high-dose PPI then swallowed topical steroids then stepped elimination diets, and the rescue path of dupilumab and graduated dilation when standard therapy fails.
Start with the disease, because every diagnostic criterion and every drug in the ladder is downstream of one mechanism. Eosinophilic esophagitis is a Th2 allergic disease of the esophageal mucosa, triggered by food allergen exposure in a genetically susceptible host, and ending in fibrosis, rings, and food impaction if it is left untreated. The cytokines do the work. Interleukin five stimulates eosinophil production in the marrow and recruits eosinophils to tissue. Interleukin four and interleukin thirteen act on the esophageal epithelium itself. They drive expression of eotaxin three, which is the chemoattractant that pulls activated eosinophils across the basement membrane and into the squamous epithelium. Interleukin thirteen also disrupts the epithelial barrier directly by downregulating filaggrin and tight-junction proteins, which increases antigen access and amplifies the Th2 response in a feed-forward loop. Mast cells contribute in parallel. They infiltrate the mucosa, they can persist after eosinophils have fallen below the histologic threshold, and they explain part of why some treated patients remain symptomatic. Chronic Th2 inflammation drives subepithelial fibrosis through TGF-beta. The fibrotic phenotype is what produces the rings, the strictures, and the food bolus impactions.
The patient is the bridge between the cytokine cascade and the endoscopy. EoE is a disease of atopic men in the third and fourth decades of life. The male predominance is roughly three to one. The typical patient has a personal or family history of asthma, allergic rhinitis, atopic dermatitis, or food allergy. The presentation in adults is dysphagia for solid foods over a long arc. The patient has had multiple food impactions in restaurants. They have eaten slowly for years, they have lubricated meat with water, they have avoided breads and steak. Those adaptive behaviors mask the disease, sometimes for a decade, before an impaction in a public place forces the encounter with an emergency department. Children present differently. Feeding intolerance, abdominal pain, vomiting, and failure to thrive dominate, because the fibrotic phenotype has not yet developed. Peripheral eosinophilia is sometimes present but is neither sensitive nor specific. EoE is most common in cold, dry climates and in white patients but occurs across all demographics.
The encounter that surfaces most diagnoses is the adult food bolus impaction, and the rule from chapter one carries directly into this material. Any adult with a food bolus impaction needs esophageal biopsies on the index endoscopy. EoE is the leading diagnosis in that scenario. The mucosa can look normal at endoscopy in five to ten percent of cases. The endoscopist who defers biopsies because the esophagus looks unremarkable will miss the diagnosis in that fraction of patients, and those are the patients who come back with another impaction.
The diagnostic threshold is at least fifteen eosinophils per high-power field on esophageal biopsy, in a patient with symptoms of esophageal dysfunction, after exclusion of other causes of esophageal eosinophilia. Three pieces of that statement carry weight. Fifteen is the number. The symptoms have to be there, because asymptomatic eosinophilia on a biopsy taken for another reason does not by itself establish the disease. And the other causes have to be excluded, which is mostly reflux esophagitis. In reflux, mucosal eosinophilia is typically less than five to ten per high-power field. The predominant inflammation is neutrophilic with reactive epithelial change, not the dense eosinophilic abscesses, basal zone hyperplasia, and spongiosis of EoE.
The biopsy strategy matters because EoE is patchy. At least six specimens are taken from at least two levels of the esophagus, proximal and distal, with most operators sampling mid esophagus as well. A single-level sample misses disease in roughly a third of cases when the distribution is segmental. The candidate who biopsies only the distal esophagus because that is where reflux disease sits will read a falsely negative biopsy in a patient who has obvious proximal involvement.
The PPI trial is no longer required to make the diagnosis. That change is recent and is worth understanding mechanistically, because the boards still test the prior framework and the candidate has to know why it shifted. Until the late twenty-tens, an eight-week PPI trial was used to separate EoE from a category called PPI-responsive esophageal eosinophilia. The idea was that a PPI responder did not have true EoE. That distinction collapsed once the clinical, endoscopic, histologic, and gene expression profiles of PPI-responsive disease were shown to be indistinguishable from those of classical EoE. The PPI response turned out not to be acid suppression. PPIs block Th2 cytokine driven eotaxin three secretion by esophageal epithelial cells. The mechanism is off-target anti-inflammatory action on the same pathway the disease runs on. So the modern framework treats PPI-responsive esophageal eosinophilia as EoE, and reclassifies the PPI as a treatment rather than a diagnostic test. The practical consequence is that a patient with at least fifteen eosinophils per high-power field and the appropriate clinical phenotype has EoE at the index endoscopy. No PPI trial, no second endoscopy to confirm.
The endoscopy is then read in a standardized language, because EoE has a characteristic visual phenotype that the boards expect a candidate to describe in shared terms. EREFS is that language. The acronym stands for Edema, Rings, Exudates, Furrows, and Strictures. Edema is pallor and loss of the normal vascular pattern of the squamous epithelium. It is graded dichotomously as absent or present. Rings are concentric circumferential mucosal folds, often called feline esophagus or trachealization. They are graded none, mild, moderate when the rings are distinct but a standard adult scope still passes, and severe when the rings prevent passage. Exudates are the white papules or plaques on the surface that correspond histologically to eosinophilic abscesses. They are graded none, mild for less than ten percent of mucosal surface involvement, or severe for more than ten percent. Furrows are vertical linear grooves running along the esophageal mucosa. They are graded none, mild without visible depth, or severe with visible mucosal depth. Strictures are graded absent or present, with the diameter at the narrowest point reported.
The reason EREFS matters beyond a label is that the five components split cleanly into two subscores that mean different things. Edema, exudates, and furrows are the inflammatory subscore. They reverse with anti-inflammatory therapy. Rings and strictures are the fibrostenotic subscore. They do not reverse with drug alone. The fibrostenotic score is what predicts long-term symptom burden, dilation need, and food impaction risk. That split is why early diagnosis matters. A patient caught in the inflammatory phase before fibrosis sets in can be managed medically for years. A patient with established rings and a narrow strictured lumen needs medicine plus mechanical work, and that hand-off is the link to section three on refractory disease.
The treatment ladder has three drugs and one dietary strategy, and the candidate who arrives expecting a hierarchy with one winner will read the data wrong. The three drugs are proton pump inhibitors, swallowed topical steroids, and dupilumab, which we cover when we get to refractory disease because that is where it sits in current practice. The dietary strategy is empiric food elimination. Selection among them is driven by adherence, allergic phenotype, and patient preference, not by superiority of one option over the others. Histologic remission, defined as fewer than fifteen eosinophils per high-power field on follow-up biopsy, is the primary endpoint for any of these therapies. Remission is reassessed at eight to twelve weeks after initiation. Symptomatic remission lags behind histologic remission and does not by itself prove control of disease. Adaptive eating can mask persistent inflammation, and the fibrotic phenotype contributes to dysphagia even in the absence of active inflammation.
PPI therapy is the most accessible first-line option because it is oral, inexpensive, and familiar. The mechanism in EoE is not acid suppression. It is the same eotaxin three blockade we just walked through. PPIs interrupt the eosinophil chemoattractant signal regardless of acid exposure. Dosing for EoE is high. Omeprazole at twenty to forty milligrams twice daily, or an equivalent of esomeprazole twenty to forty milligrams twice daily or lansoprazole thirty milligrams twice daily, for at least eight weeks before reassessment. Histologic remission is achieved in roughly thirty to fifty percent of patients on the initial eight-week course. Maintenance therapy at the lowest effective dose is required because relapse on cessation is the rule. Loss of response on long-term PPI maintenance is enriched in CYP2C19 rapid metabolizers. Dose intensification often regains response in that subgroup, which is the mechanistic explanation for why some patients need twice-daily dosing despite once-daily working initially.
Swallowed topical steroids are the most reliably effective drug class in EoE. They induce histologic remission in roughly sixty to ninety percent of patients depending on dose and formulation. The mechanism is local anti-inflammatory action on the esophageal mucosa with minimal systemic absorption. Two formulations dominate and they have two delivery tricks the candidate has to know cold. Fluticasone is delivered through a metered-dose inhaler that is sprayed into the mouth and swallowed rather than inhaled. The adult dose is eight hundred eighty micrograms twice daily, which is four puffs of two hundred twenty micrograms twice daily, for a total of one thousand seven hundred sixty micrograms per day. The patient must not breathe in during actuation. The goal is mucosal contact in the esophagus, not pulmonary deposition. The patient also must not eat or drink for thirty minutes after each dose, to maximize contact time on the mucosa. The trap here is the patient who has used the same inhaler for asthma for twenty years. They inhale by reflex during actuation. Then they wonder why a steroid that worked for their lungs is not working on their esophagus.
Budesonide is the second formulation. It is delivered as an oral viscous slurry, created by mixing a one milligram respiratory suspension with a thickening agent. Sucralose, the artificial sweetener marketed as Splenda, is the canonical thickener, and honey is the alternative. The slurry coats the esophageal mucosa as it is swallowed and the viscous vehicle prolongs contact time. The dose is one milligram twice daily in adults. A budesonide oral suspension product, Eohilia, dosed at two milligrams twice daily, is now FDA approved for adults and adolescents eleven years and older. That product removes the need for compounding. The patient rinses the mouth and avoids food and drink for thirty minutes after each dose. The major side effect of either steroid formulation is oral and esophageal candidiasis. The mouth rinse exists for exactly that reason. Candidiasis when it occurs is typically managed with a brief course of topical or systemic antifungal without discontinuation of the steroid. Adrenal suppression and growth retardation are uncommon at standard doses but should prompt consideration of dose reduction at maintenance. Symptom recurrence after stopping steroids is the rule, with roughly ninety percent of patients relapsing within a year and a mean time to recurrence of around nine months. That recurrence rate is why maintenance therapy is the default for steroid responders.
Empiric elimination diets work because EoE is fundamentally a food-allergen disease. Removal of the trigger antigens reverses the Th2 response. Three empiric diets sit on a ladder, and the remission rate falls as the diet loosens. The six-food elimination diet eliminates milk, wheat, eggs, soy, nuts, and seafood for six to eight weeks. It produces histologic remission in roughly seventy to seventy-five percent of adults. The four-food elimination diet drops nuts and seafood, eliminating milk, wheat, eggs, and soy. It produces remission in about fifty to sixty-five percent. The two-food elimination diet eliminates only milk and wheat. It produces remission in roughly forty percent.
The modern approach is step-up rather than step-down. The patient starts with a two-food elimination. They escalate to four-food and then six-food only if remission is not achieved. The cumulative endoscopic burden is lower than the older step-down strategy that began with six-food elimination. Once remission is induced on whichever protocol, foods are reintroduced one at a time with repeat endoscopy and biopsy. That sequence identifies the specific trigger or triggers, at which point the patient can return to a less restrictive diet that excludes only the trigger food.
The careful-reasoning failure with diets is the allergy test. Skin prick testing and serum IgE panels do not reliably identify EoE trigger foods, because the disease is largely IgG4 mediated rather than IgE mediated. Allergy-test directed diets achieve histologic remission rates around forty-five percent, compared with seventy to seventy-five percent for empiric six-food elimination. That difference is the reason empiric elimination is the standard. Milk is the dominant trigger, identified in roughly fifty to sixty percent of responders to six-food elimination, followed by wheat in approximately thirty to sixty percent. Soy, eggs, and nuts contribute to a smaller fraction. The elemental diet, which is an amino-acid based formula, achieves histologic remission in over ninety percent of patients. It is rarely tolerated outside the pediatric or refractory adult setting because of palatability and cost, and it is reserved for severely symptomatic or polysensitized patients who cannot identify a maintainable elimination strategy.
Selection between drug and diet is patient-driven. A strongly atopic patient with known food allergies may prefer elimination because it addresses the antigen directly. A patient who travels frequently or cannot reliably control food sources may prefer a swallowed topical steroid. A patient with concomitant typical reflux often prefers a PPI as the dual-purpose drug. The ladder allows the candidate to match the strategy to the patient rather than escalating mechanically. That principle is what the boards reward.
Now to refractory disease, because the patient who fails the ladder is the patient who needs the most careful thinking. Refractory EoE is disease that fails to achieve histologic or clinical remission after an adequate trial of first-line therapy. The modern definition increasingly includes a second group: patients who do achieve histologic remission but who have persistent dysphagia from established fibrostenotic change. Those two failure modes need different answers, and the dominant paradigm now combines them. The biologic answer addresses the inflammatory component upstream of eosinophil and mast cell infiltration. The mechanical answer addresses the fibrotic component with esophageal dilation. Either alone is incomplete in the patient who has both.
Dupilumab is the biologic answer. It is a fully human monoclonal antibody against the alpha chain of the interleukin four receptor. The mechanism is upstream and broad. IL-4R alpha is shared by both interleukin four and interleukin thirteen in their respective signaling complexes. The type one IL-4 receptor uses IL-4R alpha plus the common gamma chain and binds interleukin four. The type two IL-4 receptor uses IL-4R alpha plus IL-13R alpha one and binds both interleukin four and interleukin thirteen. Blocking IL-4R alpha simultaneously interrupts both cytokines. That collapses the eotaxin three production that drives eosinophil recruitment to the epithelium, and it reverses the IL-13 mediated barrier disruption. The drug acts upstream of every downstream eosinophil signal.
The FDA approved dupilumab for EoE in May of twenty twenty-two, in patients twelve years and older weighing at least forty kilograms. The approval was extended in January of twenty twenty-four to children one year and older weighing at least fifteen kilograms. The adult and adolescent dose is three hundred milligrams subcutaneously every week. That dose and frequency is what the LIBERTY-EoE-TREET phase three trial established. The trial enrolled patients with EoE refractory to PPI therapy and at least fifteen eosinophils per high-power field. They were randomized to weekly dupilumab, dupilumab every two weeks, or placebo, for twenty-four weeks. Both dupilumab arms used three hundred milligrams subcutaneously. Histologic remission was achieved in roughly sixty percent of patients on either dupilumab dose. Only weekly dupilumab improved dysphagia significantly compared with placebo. Every two week dosing cleared eosinophils as well as weekly dosing, but it did not improve symptoms over placebo.
That dissociation is the teaching point. The same drug at half the cumulative dose cleared the histology and failed to clear the dysphagia. The implication is that EoE dosing is weekly, full stop. The allergist who knows dupilumab for atopic dermatitis at every two weeks will reflexively start it at every two weeks. The patient will come back at three months with mildly improved symptoms and persistent eosinophils. Escalation to weekly dosing recovers the symptomatic response. The careful-reasoning failure is reading a partial histologic response on every two week dosing as treatment failure and switching to a different biologic. The right move is to give the patient the FDA-approved dose.
The other benefit of dupilumab in the polyatopic EoE patient is that it is also approved for asthma, atopic dermatitis, and chronic rhinosinusitis with nasal polyps. A patient with EoE plus atopic dermatitis plus asthma plus nasal polyps was previously looking at multiple separate biologics. Dupilumab covers all of them with a single agent. The principal side effects are injection site reactions and conjunctivitis. Other biologics under investigation include cendakimab, which targets interleukin thirteen directly. Benralizumab targets interleukin five receptor alpha and depletes eosinophils through antibody-dependent cellular cytotoxicity. Mepolizumab targets interleukin five. None of those three are FDA approved for EoE at this time. The board-relevant point is that dupilumab is the only approved biologic. It acts at IL-4R alpha to block both interleukin four and interleukin thirteen. It is dosed weekly, because every two week dosing cleared eosinophils but failed to improve dysphagia in the pivotal trial.
The mechanical answer is graduated esophageal dilation. The historical concern that EoE strictures were uniquely fragile has been substantially revised. Pooled data across more than thirteen hundred dilations show a perforation rate around two tenths of a percent. That is comparable to standard endoscopic dilation perforation rates of one to four tenths of a percent for benign strictures generally. The drift in practice is from a procedure that was deferred out of fear to one that is offered routinely when the fibrotic phenotype warrants it. The technique combines several principles. Dilation begins with a small caliber and advances gradually. The target is typically three millimeters of luminal increase per session. The goal diameter is in the fifteen to eighteen millimeter range over multiple sessions for symptom control. Either through-the-scope balloons or wire-guided Savary bougies are used. There is no clear superiority of one over the other.
The mucosal tears the older literature called complications are now understood as the mechanism of the dilation itself. Rents are seen in roughly seventy-five to eighty-five percent of EoE dilations. They are expected. They reflect the principle that the fibrotic ring is being broken open by graduated mechanical force. Patients should be counseled in advance to expect transient post-dilation chest pain. The reported rates run from a few percent up to as high as seventy percent in some series, depending on the dilator and the cohort. The pain typically resolves within hours to days. Dilation does not treat the underlying inflammation. So it must be paired with medical maintenance therapy, which is PPI or swallowed topical steroid or dupilumab or an elimination diet, to prevent re-stricturing. Dilation is also safe and effective in patients who are responsive to medical therapy but who have residual fibrotic narrowing. It should not be deferred indefinitely in those patients because of an outdated perception of high risk. The principal contraindications remain active deep ulceration, suspicion of perforation, and a high-grade stricture in a patient who has not had pretreatment to soften the inflammation.
The combined-modality approach is the dominant paradigm for refractory disease. The patient with fibrostenotic EoE who has failed PPI and topical steroids, and who has progressive dysphagia despite some inflammatory response, is a candidate for dupilumab plus serial dilation. Dupilumab controls the upstream Th2 signaling. Dilation addresses the established fibrosis. Either modality alone is incomplete.
The cleanest careful-reasoning failure on this material is the patient who has done well. The patient is in histologic remission on dupilumab. They have fewer than six eosinophils per high-power field on biopsy. They still have dysphagia, and they have a fourteen millimeter fibrotic narrowing on endoscopy. The temptation is to read that as treatment failure and switch the biologic. The right read is that the drug is doing exactly what it is supposed to do, and the dysphagia is mechanical. The patient needs dilation, not a different biologic. The diagnostic gate at the top of the episode, which is histology, separates the two failure modes. Active inflammation needs more or different anti-inflammatory therapy. Established fibrosis needs mechanical work.
Pull the episode together along the cytokine axis it began on. EoE is a Th2 allergic disease. Interleukin four and interleukin thirteen drive eotaxin three production in the epithelium, eosinophils flood the squamous mucosa, mast cells contribute in parallel, and chronic TGF-beta signaling lays down fibrosis. The diagnostic gate is fifteen eosinophils per high-power field on at least six biopsies from at least two levels, with no PPI trial required, because PPI-responsive disease is EoE and the PPI is a treatment. The endoscopy is read with EREFS, where edema and exudates and furrows are inflammatory and reverse with drug, and rings and strictures are fibrotic and need mechanical work. The treatment ladder is PPI, swallowed topical steroid, and empiric elimination diet, with the choice driven by adherence, atopic phenotype, and patient preference rather than by superiority of one option. PPI dosing is twice daily. Fluticasone is sprayed and swallowed not inhaled, eight hundred eighty micrograms twice daily, with thirty minutes of no food or drink afterward. Budesonide is a viscous slurry mixed with sucralose, one milligram twice daily, with the same post-dose interval. Step-up empiric elimination is the standard because allergy testing does not reliably identify EoE triggers. Milk is the dominant trigger. Refractory disease is met with dupilumab three hundred milligrams subcutaneously weekly, never every two weeks for EoE. The mechanical side is graduated dilation, now safer than once thought, with mucosal rents the expected mechanism rather than a complication.
The next episode covers the infectious and direct-injury esophagitides. Candida, HSV, and CMV are sorted by host context and ulcer morphology, with HSV at the epithelial edge and CMV at the stromal base. Pill esophagitis is contact injury at sites of esophageal narrowing. Caustic injury is graded by the Zargar endoscopic scale. The answer to the depth of the injury lives in the chemistry of the agent and the time it sat on the mucosa.