Eosinophilic and Infectious Esophagitis: Eosinophilic Esophagitis: Diagnosis Through Refractory Disease

0:00Welcome to Board Pearls. This is episode one of two of the Eosinophilic and Infectious Esophagitis chapter, in the Esophageal Disorders module. This episode is eosinophilic esophagitis from diagnosis through refractory disease, and every criterion and every treatment comes out of one mechanism, so we start there.

0:19Eosinophilic esophagitis is an allergic disease of the esophageal lining, driven by food allergens in a susceptible person, that ends in scarring, rings, and food getting stuck if it's left alone. The engine is a set of allergic-type cytokines. Two of them, interleukin four and interleukin thirteen, act right on the esophageal lining and drive it to make eotaxin, the signal that pulls eosinophils out of the blood and into the surface layer. Interleukin thirteen also breaks down the barrier between the lining cells, which lets more antigen in and feeds the whole process forward. Mast cells pile in alongside the eosinophils, and they can linger after the eosinophils have been driven down, which is part of why some treated patients still have symptoms. And the chronic inflammation lays down scar tissue underneath, and that scarring is what produces the rings, the strictures, and the food impactions. Hold that arc, inflammation first, scarring later, because the whole diagnostic and treatment scheme maps onto it.

1:20The typical patient connects the biology to the clinic. This is mostly atopic men in their twenties and thirties, usually with a personal or family history of asthma, hay fever, eczema, or food allergy. In adults the complaint is trouble swallowing solids over a long stretch of years, with food impactions, and the giveaway is the adaptations: eating slowly, washing meat down with water, avoiding bread and steak. Those habits hide the disease for years until an impaction in a restaurant forces the issue. Children look different, with feeding trouble, pain, vomiting, and poor growth, because the scarring hasn't developed yet. So the same disease looks inflammatory in a child and fibrotic in an adult.

2:04The encounter that surfaces most adult diagnoses is that food impaction, and the rule from chapter one carries straight over: any adult with a food impaction gets esophageal biopsies at that first endoscopy, because eosinophilic esophagitis is the leading cause and the lining can look completely normal in a small fraction of cases. Skip the biopsies because it looks unremarkable and you miss exactly the patients who come back with another impaction.

2:29The diagnosis is at least fifteen eosinophils per high-power field on biopsy, in a patient with swallowing symptoms, after excluding other causes of esophageal eosinophilia. Three parts of that carry weight. Fifteen is the number. The symptoms have to be there, because incidental eosinophils on a biopsy taken for another reason aren't the disease. And you have to exclude the other causes, mainly reflux, which typically shows fewer eosinophils and a different, neutrophil-predominant picture. Because the disease is patchy, you take at least six biopsies from at least two levels, upper and lower, since sampling only the distal esophagus will read falsely negative in someone whose disease is up high.

3:09One important shift: you no longer need a proton pump inhibitor trial to make the diagnosis, and the reasoning behind that is worth understanding. It used to be thought that a patient whose eosinophilia responded to a proton pump inhibitor had something other than true eosinophilic esophagitis. That turned out to be wrong, because the two look identical down to the gene expression, and because the drug isn't working by suppressing acid at all, it's blocking that same eotaxin signal the disease runs on. So a response to the drug doesn't rule the disease out; the drug is simply one of its treatments. The practical result is that a patient with fifteen or more eosinophils and the right clinical picture has the disease at that first endoscopy, with no drug trial and no second scope needed to confirm.

3:52You then describe the endoscopy in a standard shared language, because the disease has a characteristic look. The five features are edema, seen as pallor and loss of the normal vascular pattern; rings, the concentric circular folds sometimes called a trachea-like esophagus; exudates, the white papules that correspond to eosinophil-rich patches; furrows, the vertical linear grooves; and strictures, with the narrowest diameter recorded. And the reason this is more than a label is that those five split into two meaningful groups. Edema, exudates, and furrows are the inflammatory features, and they reverse with anti-inflammatory treatment. Rings and strictures are the scarring features, and they do not reverse with a drug. The scarring features are what predict long-term symptoms, dilation need, and impaction risk. That split is why catching the disease early matters: a patient still in the inflammatory phase can be managed with medication for years, while a patient with established rings and a narrow lumen needs medication plus mechanical work, which is the bridge to refractory disease.

4:57There are three medical treatments and one dietary strategy, and the thing to get right is that there's no single winner, so you match the option to the patient rather than ranking them. The three drugs are a proton pump inhibitor, a swallowed topical steroid, and, for refractory disease, dupilumab; the dietary strategy is empiric food elimination. For any of them, the goal is histologic remission, driving the eosinophils back below the threshold, reassessed at about two to three months, and you can't rely on symptoms alone, because adaptive eating hides ongoing inflammation and scarring causes dysphagia even when the inflammation is controlled.

5:34The proton pump inhibitor is the most accessible first option because it's oral, cheap, and familiar, and again it works here by blocking the eotaxin signal, not by suppressing acid. It's dosed high and twice daily, for at least eight weeks before you reassess, and because relapse on stopping is the rule, responders stay on a maintenance dose. Some patients who respond initially lose it over time and regain it with a higher dose, which reflects how fast they metabolize the drug.

6:03Swallowed topical steroids are the most reliably effective drug class, working locally on the lining with little absorbed, and they come in two forms with two delivery tricks you have to know cold. Fluticasone is given through a metered-dose inhaler that's sprayed into the mouth and swallowed rather than inhaled, so the patient must not breathe in during the spray, because the goal is coating the esophagus, not the lungs, and they can't eat or drink for thirty minutes afterward to keep it in contact. The classic trap is the patient who's used the same inhaler for asthma for years and inhales by reflex, then wonders why it isn't working on the esophagus. Budesonide is the other form, mixed with a thickener like sucralose or honey into a viscous slurry that coats the lining as it goes down, given twice daily with the same rinse-and-wait-thirty-minutes rule; there's also a ready-made oral suspension now that avoids compounding. The main side effect of either is oral or esophageal thrush, which is why the mouth rinse exists, and it's usually managed with a short antifungal course without stopping the steroid. And as with the proton pump inhibitor, relapse after stopping is the rule, so responders stay on maintenance.

7:09Elimination diets work because this is fundamentally a food-allergen disease, so removing the trigger reverses the inflammation. There are three empiric diets, from most to least restrictive: one removing the six main allergen groups, milk, wheat, egg, soy, nuts, and seafood; a four-food version dropping nuts and seafood; and a two-food version removing just milk and wheat. The more you eliminate, the higher the remission rate, but the modern approach is step-up, not step-down: start with the two-food elimination and escalate only if it doesn't work, which spares a lot of the cumulative scoping the old start-with-six approach required. Once you get remission, you reintroduce foods one at a time with repeat biopsies to pin down the actual trigger, and then the patient just avoids that. The key reasoning trap here is allergy testing: skin prick and blood IgE panels don't reliably find the triggers, because this disease is driven more by IgG4 than IgE, so allergy-test-directed diets underperform empiric elimination, which is exactly why empiric is the standard. Milk is the single most common trigger, then wheat. And an amino-acid-based elemental formula works in almost everyone but is rarely tolerable outside the pediatric or severely refractory setting because of taste and cost.

8:29Choosing among drug and diet is patient-driven: a strongly allergic patient with known food triggers may prefer elimination, a frequent traveler who can't control food may prefer the swallowed steroid, and someone with typical reflux too may prefer the proton pump inhibitor as a dual-purpose drug. The point is to match the strategy to the patient rather than escalate mechanically.

8:50Now refractory disease, where the thinking has to be sharpest. Failure comes in two forms, and they need different answers. One is genuine persistent inflammation, eosinophils that won't come down on first-line therapy. The other is the patient who does achieve histologic remission but still has dysphagia from established scarring. And the dominant approach now recognizes that many patients have both, so it combines an upstream drug for the inflammation with mechanical dilation for the scarring, because either alone is incomplete.

9:21Dupilumab is the drug answer, an antibody against a shared piece of the receptor for both interleukin four and interleukin thirteen, so blocking it shuts down both cytokines at once, upstream of everything, collapsing the eotaxin production that recruits eosinophils and reversing the barrier breakdown. The single most important practical point is the dosing: in eosinophilic esophagitis it's given weekly. In the trial, less frequent dosing cleared the eosinophils just as well but failed to improve the dysphagia, while only weekly dosing improved swallowing. That dissociation is the teaching point, and it's a real trap, because a clinician used to dosing this drug less often for eczema will start it that way, the patient will come back with cleared eosinophils but persistent symptoms, and the fix is simply to use the approved weekly dose, not to switch drugs. A nice bonus is that dupilumab is also approved for asthma, eczema, and nasal polyps, so one drug covers the polyatopic patient who'd otherwise need several. Its main side effects are injection-site reactions and conjunctivitis. A few other biologics targeting interleukin thirteen or interleukin five are in development but not yet approved, so dupilumab is the one to know.

10:37The mechanical answer is graduated dilation, and the old fear that these strictures were uniquely fragile has been largely overturned; the perforation risk is comparable to dilating other benign strictures. So the practice has shifted from avoiding it to offering it when the scarring warrants. The technique is to start small and advance gradually, aiming for a few millimeters of increase per session and a target diameter reached over several sessions, with either a balloon or a bougie. And the mucosal tears the old literature called complications are now understood as the mechanism itself, the scarred ring being broken open, so they're expected and patients are warned about transient post-procedure chest pain. Dilation doesn't treat the inflammation, so it has to be paired with maintenance medication to prevent re-stricturing.

11:23So the combined approach is the model for refractory fibrostenotic disease: the patient who's failed the proton pump inhibitor and topical steroid and has progressive dysphagia gets dupilumab for the upstream inflammation plus serial dilation for the established scarring. And the cleanest trap to avoid is the patient who's actually doing well, in histologic remission on dupilumab with the eosinophils cleared but still with dysphagia and a narrow fibrotic segment. That's not treatment failure and not a reason to switch biologics; the drug is doing its job and the dysphagia is mechanical, so the patient needs dilation. The biopsy is what tells you which failure you're dealing with: active inflammation needs more or different anti-inflammatory treatment, established scarring needs mechanical work.

12:07So the whole episode along the pathway it started on: this is an allergic disease where interleukin four and thirteen drive eotaxin, eosinophils flood the lining, mast cells add on, and chronic inflammation lays down scar. The diagnosis is fifteen eosinophils per high-power field on at least six biopsies from at least two levels, with no proton pump inhibitor trial needed because a response to it is treatment, not a rule-out. The endoscopy features split into inflammatory ones that reverse with drugs and scarring ones that need mechanical work. The treatments are a proton pump inhibitor, a swallowed topical steroid, and empiric elimination, chosen by adherence, allergy profile, and preference rather than ranked. Fluticasone is sprayed and swallowed, never inhaled, with no food for thirty minutes after; budesonide is a viscous slurry with the same rule. Step-up empiric elimination is standard because allergy testing doesn't find the triggers, and milk is the most common one. Refractory disease gets dupilumab dosed weekly, never less often for this disease, plus graduated dilation, which is safer than once believed, with the mucosal tears being the mechanism rather than a complication.

13:16The next episode covers the infectious and direct-injury esophagitides. Candida, herpes, and CMV are told apart by the patient's immune status and the shape of the ulcer, with herpes at the surface edge and CMV at the deep base. Pill esophagitis is contact injury at the spots where the esophagus narrows. And caustic injury is graded endoscopically, with the depth of the damage coming down to the chemistry of the agent and how long it sat on the lining.

13:44For the full chapter, the practice vignettes, and the topic-tagged question bank, head to board pearls dot com. You'll find the rest of the series on Apple Podcasts, Spotify, or wherever you listen to podcasts. That brings us to the end of episode one of two of chapter five, and I'll see you in the next one.