Approaches and Challenges to Management of Pediatric and Adult Patients With Eosinophilic Esophagitis

Abstract

Treatment of eosinophilic esophagitis has progressed from elemental formula for children and esophageal dilation for adults to selective exclusion of food triggers and swallowed topical corticosteroids. Management guidelines are available from the American Gastroenterological Association and the Joint Task Force on Allergy Immunology Practice Parameters. We cannot, however, evaluate the efficacy of treatments without a definition of response. We propose a treat-to-target approach, based on symptoms and findings from endoscopy and histology. This approach addresses dissociations between outcomes, such as symptom persistence despite normalization of histologic features and symptom resolution after esophageal dilation despite histologic features of active disease. Eosinophilic esophagitis can now be treated with biologic agents that target specific immune pathways, and findings from prospective trials have indicated that less-restrictive, empiric, elimination diets can be effective and reduce the need for repeated endoscopic assessment of disease activity during food reintroduction. We also discuss eosinophilic esophagitis subtypes, factors associated with disease, and advances in management.

During the last 3 decades, care of patients with eosinophilic esophagitis (EoE) has progressed from esophageal dilation for adults and elemental formula for children to selective exclusion of common food triggers and swallowed topical corticosteroids. In the decade since the initial recommendations for diagnosis of EoE were published, guidelines for clinical care and criteria for conduct of research studies have been developed.^1–5 Although no treatments for EoE have been approved by the Food and Drug Administration (FDA), phase 2 and 3 trials of esophageal-optimized corticosteroid preparations and eosinophil-targeted therapeutic agents are underway. We review the goals and rationales for treatment, differences in management of children vs adults, new medical and dietary approaches, special-case scenarios, and considerations for management of patients with EoE.

Treatment of Eosinophilic Esophagitis

Studies of EoE progression and placebo-controlled trials have provided rationale for the importance of treating EoE. EoE is treated with dietary approaches (eg, eliminating specific allergens that cause EoE) and pharmacologic agents (eg, eliminating eosinophils directly or interrupting the signaling pathways that perpetuate inflammation) (see Figure 1).^1,2 If left untreated, EoE leads to complications that are primarily related to esophageal remodeling.^6,7 Clinical and genetic analyses have revealed that EoE, like other chronic inflammatory diseases, has multiple phenotypes. In some patients with EoE treated with elimination or elemental diets, persistent inflammation and evidence for aeroallergen-induced sensitization and concurrent atopic disease indicate a contribution of nonfood-based antigens.⁸ Mechanistic studies have provided information about the inflammatory pathways that contribute to EoE9 and identified targets for therapy, as opposed to the broad anti-inflammatory effects of corticosteroids. EoE and its treatment can place a great burden on patients, so it is important to balance cost, quality of life, and long-term effect of treatment to optimize patient care.

Figure 1.

Pathogenesis and therapeutic targets for EoE. Pathogenesis of EoE involves presentation of dietary and environmental antigens to T cells leading to an inflammatory response mediated by T-helper 2 cells and IL13. In the esophageal mucosa, IL5 and eotaxin-3 recruit eosinophils from the bone marrow. Eosinophils and mast cells secrete proteases, cytokines, and histamines to promote inflammation and tissue remodeling. Phase 2 trials of patients with eosinophilic gastrointestinal disorders have studied the efficacy of monoclonal antibodies against IgE, IL4 receptor–α, IL5, IL5 receptor, IL13, and sialic acid binding Ig-like lectin 8 (Siglec8), as well as an oral antagonist of prostaglandin D2 receptor 2. Other targets for treatment could include IL9, IL15, thymic stromal lymphopoietin (TSLP), transforming growth factor–β (TGF-β), and eotaxin 3.

The main goals of therapy are elimination of symptoms, reductions in mucosal inflammation, restitution of disease consequences, and prevention of additional complications.^10,11 There are challenges to achieving these goals. First, it is a challenge to assess symptoms in patients related to EoE; only recently have validated measures of patient-reported outcomes become available. Most problems in patients with EoE are related to longstanding dysphagia—patients often develop coping mechanisms to facilitate passage of food. The acronym ‘IMPACT’ identifies these adaptive behaviors: Imbibe fluids with meals, Modify food (cutting into small pieces, pureeing), Prolong meal times, Avoid hard texture foods, Chew excessively, Turn away tablets/pills. Identification of patients’ most prominent symptoms allows researchers to track the effects of treatment. This is a challenge for trials in which study participants may have diverse symptoms—homogenous entry criteria are required to document symptom outcomes. Second, assessment of the esophageal mucosal surface requires endoscopic procurement of biopsies. Anesthesia, required time away from work and school, complications, cost, and effects on quality of life can interfere with repeated endoscopy. Furthermore, biopsies represent only a small part of the esophageal mucosal surface. Third, the optimal time to assess the effects of treatment is difficult to identify. Because symptoms do not always correspond with mucosal inflammation, they cannot be used reliably to determine mucosal healing.¹² Duration of treatment assessment is expected to vary based on the time course required for resolution of inflammation with specific treatments.

Studies of treatment effects differ between children and adults. Children’s eating behaviors change with time, so questions must change with age (see Table 1). There is evidence that the developing brain is affected by repeated exposure to anesthesia, so the number of endoscopies performed and age of patients should be considered.¹³ Because EoE is a chronic disease, long-term approaches for children, as they relate to inducing disease remission and quality of life, are important to consider. Social anxiety around meals can be related to disease activity but also treatments with elimination diets.¹⁴ Growth and development are priorities in care of pediatric patients. Not only should patients continue to follow a healthy growth curve, but they also must continue to meet development milestones that include eating skills. Selection of therapy must carefully balance the burdens of symptoms, repeated monitoring of disease activity by endoscopy, and daily use of a medical or dietary treatment.

Table 1.

Eating Behaviors in Pediatric Patients With Eosinophilic Esophagitis

Variable	Infants or toddlers	Grade school	Adolescents
Duration of meals	Mealtimes longer than sibling or rest of family; often leaves and comes back to the table; grazes on small volumes of liquid or food	Mealtimes longer than friends; returns from school with full lunchbox	Avoids social dining due to prolonged mealtime or fear of food getting stuck
Coping behaviors	Preference for liquids and soft foods over solid foods	Use of large amounts of dips, sauces, or liquids to help swallowing; may have narrow range of preferred foods	Always needs water bottle or liquids with meals
	Pockets food in cheek for prolonged periods and/or spits food out; dips foods in liquids	Prolonged chewing of food before swallowing	Prefers a soft-textured diet
Food selection	Difficulty advancing diet from pureed baby food; demonstrates feeding refusal or fussy behavior during meals	Difficulty to refusal to expand diet with new flavors, types of foods, or textures	Avoidance of certain food textures, specifically meats, bread, rice, raw fruits, and vegetables

Analyses of esophageal biopsies have come under increasing scrutiny because histologic features of EoE vary within and among patients. There is little correlation between symptom severity, such as dysphagia, and eosinophil density.¹² Despite these issues, no other activity measure or biomarker has been identified to replace histology. Assessing histologic response to therapy relies on counting eosinophils, with a number of different thresholds, including a percent decrease or a specific number (<15, <6, or 0). Assessment strategies are being developed that include not only eosinophil count, but also histologic feature, incorporating multiple aspects of mucosal inflammation.¹⁵ Collins et al¹⁶ developed the EoE histologic scoring system based on readily identified features in biopsies, including basal zone hyperplasia, dilation of intercellular spaces, and lamina propria fibrosis. This system assesses the severity and extent of inflammation. EoE histologic scoring system scores are increasingly included as secondary outcomes in trials of agents for treatment of EoE.

Defining Response to Treatment

There is no standardized definition for response to therapy for EoE.^10,11 For trials that require regulatory approval, the FDA has proposed the co-primary end points of histopathology features (peak esophageal eosinophil density) and symptom-based patient-reported outcomes.¹⁷ Endoscopic activity has been used in trials as a secondary end point to determine the efficacy of therapeutic agents. In addition to assessing mucosal inflammation, endoscopy can be used to identify esophageal remodeling (development of rings and strictures), which cannot be identified in mucosal biopsies and has been associated with symptoms.^18,19 The esophagus can be imaged using barium esophagrams and impedance planimetry (with the functional lumen imaging probe)—these are probably more accurate in quantification of esophageal remodeling. An effective treatment should improve symptoms and reduce histologic and endoscopic features of disease (Figure 2). Treatment of symptoms sounds like a good goal but can be problematic because symptoms do not always associate with esophageal remodeling or pathology features of EoE. Histologic features indicate disease pathogenesis and have been associated, in retrospective studies, with disease complications. Finally, endoscopic demonstration of clinically relevant strictures should raise concerns even in the absence of symptoms.

Figure 2.

Goals of management. Goals of treatment are to minimize symptoms and endoscopic histopathologic features of EoE.²⁰ Maximal symptoms are dysphagia with every meal, despite a modified, soft-solid diet, whereas minimal symptoms are complete absence of dysphagia without any dietary avoidance based on food texture. Minimal histologic activity is defined as 0 eos/hpf. Minimal endoscopy activity is defined as an absence of endoscopic features of inflammation (furrows, exudate, or edema) and an esophageal diameter >20 mm. Effective treatment would normalize all of these disease features (white center); whereas ineffective therapy would result in maximal activity of these features (black outer circle). The composite approach identifies clinically relevant issues with dissociation among outcomes. For example, anti-inflammatory therapies often normalize histologic features of EoE, but have small effects on symptoms generated by fibrostenotic esophageal strictures (yellow circle 1). Esophageal dilation, however, is effective at relieving symptoms of dysphagia and endoscopic stricture, but does not reduce mucosal inflammation (yellow circle 2).

Reviews of patients’ symptoms and pathologic and endoscopic features can result in multiple scenarios (see Table 2). Patients’ disease might seem to be refractory to a specific therapy, based on their symptoms of pathologic and endoscopic features. Whereas a complete responder would have resolution of all 3 of these features, a definite nonresponder would have persistent activity of all 3. Assessment of permutations of response to 1 feature that is dissociated from the other features is a challenge for physicians. For example, improvements in symptoms and endoscopic features of disease after esophageal dilation occur without changes in histologic features. Conversely, resolution of histologic features after treatment with swallowed topical corticosteroids or elimination diets do not always produce improvement in endoscopic remodeling features, esophageal strictures, or associated dysphagia.

Table 2.

Targeted Therapies for Eosinophilic Esophagitis

Variable	Symptom responsea	Symptom nonresponse
Histologic responseb
Endoscopic responsec	Successful treatment outcome Consider maintenance therapy and clinical and endoscopic follow-up with interval based on severity of presentation	Missed esophageal stricture or possible subepithelial or muscular eosinophilic inflammation; consider barium esophagram with pill, empiric esophageal dilation, esophageal manometry, or impedance planimetry
		Functional dysphagia
		Non-EoE cause of dysphagia, including esophageal Candidiasis and esophageal dysmotility
Endoscopic nonresponse	Verify that patient is not using adaptive eating behaviors or food avoidance to minimize symptom reporting (see IMPACT behaviors in text)	Perform dilation of esophageal stricture Possible subepithelial or muscular eosinophilic inflammation
	Consider esophageal dilation for strictures <15 mm to reduce risk for future food impaction
Histologic nonresponse
Endoscopic responsec	After esophageal dilation in absence of effective medical or diet therapy	Symptoms may be related to active esophageal inflammation
	Consider alternative treatment approach to achieve histologic response Consider clinical follow-up for concerns of disease progression in setting of ongoing inflammation	Possible missed esophageal stricture; consider barium esophagram with pill, empiric esophageal dilation, or impedance planimetry Possible esophageal dysmotility due to EoE
Endoscopic nonresponse	Verify that patients are not using adaptive eating behaviors and/or food avoidance to minimize symptoms (see IMPACT behaviors in text)	Complete nonresponder; consider nonadherence or additional food allergen (for patients on dietary therapies)
	Consider alternative treatments to produce change in histologic features	Suboptimal dosing or delivery of topical corticosteroids
	For patients who refuse alternative treatments, consider follow-up examinations, with interval based on severity of initial presentation	Consider esophageal dilation to reduce symptoms while waiting for new therapies
		Consider use of combination therapy, clinical trials of therapeutic agents, or off-label use of biologic agents, if patients have indications for non-EoE atopic conditions

NOTE. Bold font indicates primary considerations for each scenario.

^aSymptom response is defined by absence of dysphagia without the need for dietary modification and/or avoidance of harder-texture foods.

^bHistologic response is defined by <15 eos/hpf.

^cUnless specified, endoscopic response means improvement of esophageal remodeling features (ring severity <2 and diameter >15 mm) and reductions in inflammatory features (edema, exudate, furrows).

Thresholds of response for each of these outcomes are not well-defined. In practice, resolution of dysphagia in the absence of diet modification and avoidance based on food texture should be the goal.²⁰ Normalization of the esophageal epithelium is an ideal target (such as 0 eosinophils/high-power field [eos/hpf]), but improvements in outcome have been associated with <15 eos/hpf.²¹ Incorporation of additional markers of esophageal inflammation and tissue injury that are part of the EoE histologic scoring system will likely improve histologic assessment.¹⁶Reductions in endoscopic features, of disease, such as inflammatory features of exudate, furrows, or edema, combined with maintenance of an esophageal diameter ≥16 mm, are treatment goals.

Optimizing Treatment

Proton Pump Inhibitors

The European Guideline on EoE and the AGREE (A working Group on PPI Responsive Esophageal Eosinophilia) consensus statement state that proton pump inhibitor (PPI) therapy is no longer necessary for diagnosis of EoE and positioned PPIs as a therapeutic option for EoE.^4,22,23 The recommendation effectively eliminated the terminology of PPI-responsive esophageal eosinophilia. The American Gastroenterological Association Joint Task Force on Allergy and Immunology (AGA-JTF) guidelines for EoE recommend PPI therapy for esophageal eosinophilia, based on reports of reductions in histologic features of disease from 42% in observational studies, although there is heterogeneity in responses to PPIs.^5,24 Although response to PPIs is lower than reported with most corticosteroid studies, their safety and ease of administration can make them a good first-line treatment option—especially given the lack of FDA-approved formulations of swallowed topical corticosteroids that are optimized for esophageal delivery. Budesonide is available in a tablet formulation that was recently approved by the European Medicines Agency.²⁵

Although the removal of the PPI requirement for the diagnosis of EoE simplified the diagnostic criteria, it led to questions about management of EoE. Trials conducted in patients with EoE usually excluded patients with PPI-responsive esophageal eosinophilia, so it is unclear whether their results can be extrapolated to treatment-naïve patients who respond to PPIs. Studies are needed to determine the effects of combining PPI therapy with topical corticosteroids or elimination diets. Moreover, increasing our understanding of the mechanisms of response to PPIs and factors associated with it could optimize management of patients. It is important to learn more about the long-term outcomes of patients with esophageal eosinophilia and EoE who respond to PPIs—these have been examined in a few studies in children and adults.^26–28 A prospective, uncontrolled case series reported that most patients with EoE who respond to initial, high-dose PPI therapy maintain their response upon dose reduction.²⁷ Case reports, however, have noted loss of initial PPI response during long-term follow-up.²⁹ Polymorphisms in CYP2C19 that result in hypermetabolism of PPIs have been associated with loss of initial response to these drugs.²⁶

Swallowed Topical Corticosteroids

Swallowed topical corticosteroids are a popular primary therapy for EoE and were the only therapy in the AGA-JTF guideline to receive a strong recommendation based on moderate-quality evidence.⁵ Overall, about two-thirds of patients enrolled in placebo-controlled trials demonstrated a histologic response.²⁴ The AGA-JTF technical report identified heterogeneity in the histopathologic response rate to steroids in published clinical trials, demonstrating nonresponse to topical corticosteroids in proportion of patients ranging from 5% to 69%.²⁴ Current drawbacks to corticosteroids include the lack of an FDA-approved corticosteroid for EoE and absence of readily available formulations optimized for esophageal delivery in the United States. Insurance coverage for the repurposed use of asthma topical corticosteroids is a common cost concern. The European Medicines Agency recently approved a budesonide tablet for EoE, and a liquid suspension of budesonide is currently in phase 3 trial in the United States.^25,30 Studies have reported factors associated with nonresponse to steroids, including older age,³¹ atopic status,³¹ absence of a CC polymorphism in the transforming growth factor β1 gene (TGFB1),³² severe esophageal stricture,³³ prior esophageal dilation,³⁴ family history of EoE,³⁵ and lower dose of swallowed topical corticosteroid.³⁵ Adverse side effects have been limited to oropharyngeal and esophageal Candida infections, but systemic effects including adrenal insufficiency and bone density are being carefully evaluated. Candida infections have been largely asymptomatic and incidental findings at the time of systematic follow-up protocols in clinical trials. Reports of effects on adrenal function have been reported in observational studies using varying definitions of adrenal suppression. The known safety profile of topical steroids used for allergic rhinitis and asthma combined with the lower bioavailability for swallowed compared to inhaled steroids provide reassurance while awaiting prospectively collected data.

Elimination Diets

Diet therapy is a first-line option for children and adults with EoE. Challenges to long-term implementation and adherence to elemental formula diets led to the application of allergy test-directed elimination diets. However, high rates of false-positive and false-negative results from IgE tests for EoE caused physicians to turn to empiric, rather than test-directed, elimination diets. The six-food elimination diet (SFED) has been the best-studied diet and has an overall response rate of 60%–70%.^24,36 Although elemental, test-directed elimination, and empiric elimination diets are all recommended for treatment of EoE in the AGA-JTF guidelines, they were each assigned a conditional recommendation, due to very-low- or low-quality evidence.⁵ Reported heterogeneity in response rates could arise from confounding factors, such as patient selection or absence of consistent protocols (including counseling by dietitians, timing of biopsy collection, concomitant PPI use, or inadvertent contamination of diets).

A major limitation to the use of elimination diets in treatment of EoE is the need for repeated endoscopic biopsy assessment during the food reintroduction process. Studies have addressed this concern by reducing the number of foods that are eliminated from 6 to 10 down to 1 to 4.²⁴ Although the response rates from a lower number of eliminated foods is less than reported with an SFED approach, fewer endoscopies are required. A multicenter prospective trial in Spain tested a step-up approach in which patients initially eliminate the 2 most common food triggers (wheat and milk); patients who still have histologic features of active disease then eliminate 4 foods (milk, wheat, legumes, and eggs).³⁷ Patients who still have histologic features of active disease after elimination of these 4 foods then move on to the SFED. Using this step-up protocol, the investigators reduced the use of endoscopy of EoE for 20% of patients. A cost-effectiveness study found the step-up approach to be more cost-efficient than the SFED.³⁸ A recent randomized trial of diet therapies for pediatric patients with EoE compared a milk-elimination diet to a 4-food elimination diet. Interestingly, although symptom improvement was greater with the 4-food elimination diet, histologic features of EoE were reduced in 40% of patients in each group.³⁹

Noninvasive, office-based tests can be used to assess esophageal inflammation during diet treatments without the need for an endoscopic biopsy.⁴⁰ At this time, serologic, fecal, and exhaled breath biomarkers do not correlate with disease activity. Numbers of eosinophils in peripheral blood associate with histologic features of EoE, but detect this disease with low levels of sensitivity and specificity. Collection of epithelial cells or proteins involved in eosinophil degranulation with a sponge or string provide samples that can be analyzed with a high degree of accuracy compared with endoscopic biopsies. Transnasal endoscopy is being evaluated in validation studies and offers advantages of biopsy sampling without the need for conscious sedation or systemic anesthesia. Once validated, these technologies could support the use of dietary elimination therapies for patients who wish to avoid pharmacologic therapy.

While we wait for more accurate, less-invasive tests, less-restrictive elimination diets might be more feasible for patients with EoE. Elimination diets can be selected based on patient preference. Patients who are highly motivated to try dietary therapy might accept the high burden of repeated endoscopies to maximize overall likelihood of response to the SFED. Alternatively, eliminating 1, 2, or 4 foods might appeal to patients who want to try dietary changes before they commit to an SFED. Long-term adherence to diet elimination could be increased by permitting intermittent or subthreshold dietary exposures or diet holidays, during which patients can ingest known diet triggers while taking short courses of topical steroids. A recently completed, prospective trial by the Consortium of Eosinophilic Gastrointestinal Disease Researchers that compared milk elimination to SFED in adults with EoE will provide more information about selection of diet therapy (ClinicalTrials.gov ID: NCT02778867).

Choosing Initial Therapy for Eosinophilic Esophagitis

Selection of initial therapy for EoE involves consideration of reported efficacy, patient preferences, disease severity, availability of resources for elimination diet therapy, and practicalities of insurance coverage for off-label use of medications.²⁰ Patients and their physicians should select the treatment together. Patients who do not want to swallow topical corticosteroids each day, who are concerned about adverse effects of PPIs, or who are inclined to use alternative medical approaches, might be motivated to try an elimination diet. Patients with severe esophageal strictures <10 mm would benefit from initial esophageal dilation combined with medical or dietary therapy. Diet therapy can be selected based on practice patterns; anticipated effectiveness; long-term adherence concerns; and resources, such as allergists and dietitians. Insurance coverage for fluticasone and budesonide varies, so these drugs can be costly to patients. Similarly, limitations to availability and insurance coverage for endoscopic procedures can be on obstacle for patients who undergo repeated endoscopies during food reintroductions. Participation in clinical trials offers the opportunity to aid in development of new therapies and the practical benefits of coverage of medication and procedural costs.

Combination Therapy

Prospective studies have not evaluated the efficacy of combination therapies for EoE. Combined use of PPI therapy and swallowed topical corticosteroids seems logical for patients who have gastroesophageal reflux disease and EoE. Patients with partial but incomplete symptom and histologic responses to PPI (such as improvement in dysphagia and reduction in pathology from 75 to 25 eos/hpf) might benefit from initiation of swallowed topical corticosteroids with continued PPI therapy. Esophageal dilation is a logical addition for patients with persistent strictures after medical or diet therapy. In contrast, the combination of swallowed topical corticosteroids and elimination diet therapy has practical limitations and there are no data to support this approach. The high degree of efficacy of corticosteroids will confound attempts to monitor esophageal eosinophil density during food reintroduction and prevent identification of food triggers.

New Treatments

Therapies have been designed to target specific immune response pathways associated with development of EoE (Table 3, Figure 1).⁴¹ The AGA-JTF guidelines do not provide a recommendation for treatment of EoE with antibodies against interleukin (IL) 5, IL4, or IL13.^5,24 Three randomized, placebo-controlled trials of antibodies against IL5 reported significant reductions in histologic features of EoE.^5,24 However, these trials found no effects on symptoms beyond those of placebo, but did not include validated patient-reported outcome instruments. A double-blind, placebo-controlled trial of anti-IL13 therapy in adults reported significant reductions in histopathologic and endoscopic features of EoE.⁴² Although the study was not powered to demonstrate symptom improvement, a trend of reduction in dysphagia was reported, especially in the 50% of patients identified as steroid-unresponsive. A double-blind, randomized, placebo-controlled trial of an antibody against IL4 receptor–α found that this agent reduced patient symptoms and histologic and endoscopic features of EoE.⁴¹

Table 3.

Targets of Treatment for Eosinophilic Esophagitis and Eosinophilic Gastrointestinal Disease

Target	Role in pathogenesis	Clinical trials in EGIDs	Clinical trials in non-EGIDs
IL5 and IL5 receptor (IL5R)	Activation and recruitment of eosinophils	Mepolizumab (phase 2, pediatric trial, adult trial), reslizumab (phase 2, pediatric trial)	Mepolizumab (asthma, hypereosinophilic syndrome), reslizumab (asthma)
IL5R	Eosinophil recruitment and survival	Benralizumab (phase 2 trial)	Benralizumab (asthma, hypereosinophilic syndrome)
IL9	Promotes mast cell expansion; mast cell expression of TGFB, IL5, and IL13, and increases B-cell production of IgE	—	—
IL13	Promotes recruitment of eosinophils, barrier dysfunction, remodeling	QAX576 (phase 2, adult trial), RPC4046 (phase 2, adult trial)	Tralokinumab (asthma, atopic dermatitis)
IL4 receptor–α	Maintenance of T-helper 2 cell inflammatory process	Dupilumab (phase 2 and 3 trials of adults)	Dupilumab (atopic dermatitis, asthma, chronic rhinosinusitis with nasal polyposis), AMG 317
Prostaglandin D2 receptor 2	Recruiting T cells	OC000459 (phase 2, adult)	AZD1981, OC459, QAV680
IL15	Prevents eosinophil apoptosis, stimulates production of C-C motif chemokine receptor 3 (CCR3 or eotaxin 3), homeostasis of natural killer cells	CALY-002 (preclinical studies only)	—
SIGLEC8	Induction of eosinophil cell death Inhibition of mast cell activation	AK002 (phase 2 adult trial in eosinophilic gastroenteritis with or without EoE)	AK002 (mastocytosis, urticaria)
TSLP	Recruitment of basophils, stimulates production of IL4 to promote T-helper 2 cell responses	—	Rezepelumab (AMG 157) (atopic dermatitis, asthma)
IgE	Activation of mast cells	Omalizumab (phase 2 trial of adults)	Omalizumab (asthma, urticaria)
Integrin α4β7	Recruitment of T cells, eosinophils, and mast cells	Case reports	Vedolizumab (inflammatory bowel disease)
TNF	Synergistic effects of IL4-induced production of eotaxin 3	Infliximab (case series)	Infliximab (rheumatoid arthritis, Crohn’s disease, ulcerative colitis, psoriasis, uveitis, ankylosing spondylitis, asthma), adalimumab (Crohn’s disease, ulcerative colitis, psoriasis, rheumatoid arthritis, ankylosing spondylitis), certolizumab (Crohn’s, rheumatoid arthritis, psoriatic arthritis, spondylarthritis), golimumab (rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis, ulcerative colitis)
Eotaxin 1, eotaxin 2, eotaxin 3	Recruit eosinophils	—	GW766994
TGFB1	Increase collagen production to promote fibrosis, promote smooth muscle contraction, and worsen barrier integrity	Losartan (open-label prospective, pediatric trial)	Fresolimumab, losartan

NOTE. For more data, see references 5, 24, and 41.

EGID, eosinophilic gastrointestinal disease.

Data from studies of targeted biologic therapies for EoE therefore appear promising. An increasingly recognized proportion of patients with EoE have limited symptoms and histologic responses to induction or maintenance with topical corticosteroids and elimination diets, highlighting an unmet clinical need. Patients with EoE typically have extraesophageal allergic diseases, so systemic therapy with biologic agents could limit the need for multiple, locally active medical therapies and allergen avoidance. Furthermore, biologic therapies offer conceptual advantages of systemic therapy over topical therapy that may address the subepithelial inflammatory and remodeling aspects of EoE.

Maintenance Therapy

Arguments against maintenance treatment in EoE include the following: patients can outgrow the disease; progressive remodeling does not appear consistently, so not all patients develop complications of food impactions or strictures; chronic treatment may not be cost-effective; use of steroids can have long-term side effects; adverse effects on quality of life by dietary restrictions imposed by an elimination diet; and targets for chronic treatments are not well-defined.^7,43 Although these are reasonable arguments, evidence supports the importance of long-term treatment. As the prevalence of EoE continues to increase, one study found that only 0.5% of patients (9 of 1812) develop tolerance to previously identified food allergens that initiate EoE responses.⁴⁴ However, it is not clear whether these are patients with EoE or something else, such as gastroesophageal reflux disease. Even though EoE phenotypes are increasingly recognized, we do not have the ability to select therapy based on biomarkers. Although some patients do not develop strictures or food impaction, we cannot identify patients who will. Studies of untreated patients are likely to remain retrospective, and it will remain difficult to identify if and when progression from an exudative mucosa to a remodeled mucosa with rings or strictures occurs. Some studies have been analyzing costs and quality of life of patients with EoE, but modeling has yet to ascertain the long-term effects of treatment vs no treatment on these outcomes. Studies to identify new therapeutic targets and validate biomarkers of chronic remodeling are underway. Although EoE is not a severe chronic disease like diabetes or hypertension, long-standing damage to the esophagus can have consequences, such as esophageal stricture; food impactions; and, in young children, feeding dysfunction and malnutrition. In this regard, and because current treatments are thought to have minimal side effects, a strong case can be made to continue treatment on a long-term basis. The AGA-JTF guidelines recommend continued use of swallowed topical corticosteroids by patients in remission after short-term therapy.⁵ The recommendation acknowledged the few studies and limited evidence to support the guidance and to define what constitutes effective maintenance therapy in EoE.²⁴

In a small trial, patients were randomly assigned to groups that received low-dose budesonide (0.25 mg twice daily) or placebo for 1 year. Although the patients given budesonide had significant reductions in eosinophil density, compared with placebo, only 36% of patients maintained an eosinophil density <5 eos/hpf at 1 year. No dose-finding study has supported the choice of 0.25 mg twice daily as appropriate or sufficient vs other amounts.⁴⁵ A low maintenance dose of budesonide, compared with an induction dose of 1 mg twice daily, likely reduced the efficacy, although development of steroid-tolerance or selection of steroid-refractory patients is plausible. A large randomized controlled trial of an orodispersible budesonide tablet found that after successful induction therapy with budesonide (1 mg twice daily), >70% of adults with EoE maintained a combined clinical and pathologic (<15 eos/hpf) response at 48 weeks with either 0.5 mg or 1 mg twice-daily doses.⁴⁶ In addition, single-group, observational studies of PPIs reported sustained histologic responses in most adults, despite dose reduction.²⁴ There are few data on the long-term effectiveness of elimination diets.

Although there are pathology and symptom data to support effectiveness of prolonged medical treatment of EoE, there is uncertainty about the durability, safety, and generalizability of maintenance therapy. To reflect the limitations to the existing evidence, the AGA-JTF guideline included a qualification that patient preference be included in decisions for long-term therapy. Patients wanting to avoid the inconvenience and possible long-term adverse effects of continued therapies might choose to withhold treatment, with the understanding that undesirable outcomes of dysphagia, food impaction, feeding dysfunction, and esophageal stricture may result. In such cases, clinical follow-up is advisable.

Esophageal Dilation

Esophageal dilation was the first reported therapy for adults with EoE and is an effective strategy for management of symptoms of dysphagia resulting from strictures associated with EoE.⁵ Initial concerns for high risk of complications of esophageal perforation have not been supported in more recent, retrospective series studies. However, these were large case series studies from esophageal centers that have adopted a cautious approach to dilation in EoE. The practical application of a conservative dilation approach has been detailed in recent articles.^20,47

There are at least 3 groups of patients who might respond to treatment with dilation. Patients with strictures identified endoscopically with symptoms of dysphagia clearly benefit. Another group is patients who present with symptoms of dysphagia in the absence of endoscopically apparent stricture. Because EoE-related strictures can extend longitudinally, they may be under-recognized on endoscopy and only evident upon radiographic imaging or impedance planimetry. In the absence of imaging, cautious and empiric esophageal dilation might identify clinically relevant strictures. Finally, dilation might aide patients who deny symptoms of dysphagia because of adaptive eating behaviors and food avoidance, despite presence of high-grade esophageal strictures (<10 mm). These patients could benefit from esophageal dilation, which would reduce risk of future food impaction, even in the absence of symptoms and histopathology features.

An important yet unanswered question relates to the timing of esophageal dilation in patients with EoE. In patients with erosive esophagitis and a peptic stricture from gastroesophageal reflux disease, esophageal dilation is typically postponed to allow for mucosal healing with acid-suppressing therapy. By analogy, the significant symptom and endoscopic benefits of medical therapies in placebo-controlled clinical trials is convincing and supports the initial use of medical therapy before esophageal dilation. Studies showing improvement in esophageal distensibility with medical therapy add credence to the ability of such therapies to increase esophageal diameter in the absence of endoscopic dilation.⁴⁸ Until data from prospective studies are available, physicians will have to use their best judgment. If a patient has a high-grade esophageal stricture, prolonged food impaction, or is unlikely to adhere to medical therapy, dilation before medical therapy is reasonable. Conversely, dilation could be postponed in patients with mild-to-moderate strictures with stricture reassessment at the time of a follow-up endoscopy to evaluate the response to anti-inflammatory therapy. The effectiveness of medical therapy in improving esophageal strictures in EoE is expected to reduce the need for dilation in a subset of patients.

Special Situations

There is a growing number of subsets of patients with EoE (Table 4). Although management of EoE may be similar in most, specific scenarios warrant modification of standard of care approaches in others. Some patients may have manifestations of EoE in the setting of more generalized eosinophilic gastrointestinal disorders, such as eosinophilic gastritis; enteritis; and colitis, and may require use of systemic medical or diet therapy rather than esophageal-targeted, swallowed topical corticosteroids. Other patients may have esophageal eosinophilia as a manifestation of another underlying diseases. In these circumstances, clinical judgment that evaluates symptoms, laboratory values, and other testing will be critical. For instance, studies found a 3- to 6-fold increase in the prevalence of inflammatory bowel diseases in patients with EoE. Although esophageal Crohn’s disease exists, it is rare, with features distinct from EoE.⁴⁹ Systemic therapies for inflammatory bowel diseases might also reduce symptoms or features of EoE. Recognition of concomitant functional esophageal disorders in patients with EoE is important to avoid unnecessary escalation of therapy based on symptoms in the absence of objective evidence of disease activity.

Table 4.

Special Situations

Situation	Comments
EoE with GERD	The population prevalence of GERD makes overlap of GERD and EoE inevitable. The presence of typical GERD symptoms and manifestations (erosive esophagitis, Barrett’s esophagus) make concomitant therapy for GERD appropriate in such patients.
EoE with concomitant other EGIDs	Current diagnostic criteria exclude patients with eosinophilic gastritis, enteritis, or colitis from a formal diagnosis of EoE. An increasingly recognized proportion of patients with other EGIDs, however, have concomitant findings of EoE. Identification of generalized EGID in addition to esophageal involvement may alter treatment selection.
EoE with IgE-mediated food allergy	Most patients with EoE also have atopic diseases that could include IgE-mediated food allergy, such as oral allergy syndrome and food-associated anaphylaxis. Identification of concomitant food allergy will affect the management of dietary elimination for EoE.
EoE in asymptomatic subjects	Diagnosis of EoE requires esophageal eosinophilia and symptoms of esophageal dysfunction. Patients with early-stage disease, different phenotypes of disease, or compensated eating habits can have minimal symptoms. The presence of endoscopic features of EoE would support the diagnosis of EoE in such cases. Proactive treatment of such individuals to prevent disease progression vs a watch-and-wait strategy are considerations.
Oral immunotherapy-induced EoE	Oral immunotherapy is being used to induce immune tolerance to food allergy. EoE reported in 5.3% of patients with upper endoscopy with biopsy performed to evaluate gastrointestinal symptoms during immunotherapy for food allergy. Most patients with immunotherapy-induced EoE have been managed by stopping this treatment.50
EoE with celiac disease	Co-existent EoE and celiac disease reported in pediatric and adult case series, but an association between diseases is unclear.51 The diseases have distinct pathogenetic mechanisms but share certain demographic risk factors. Wheat is only one of multiple potential food triggers for EoE.
EoE with IBD	A 5- to 7-fold increased prevalence of EoE identified in patients with IBD (ulcerative colitis or Crohn’s disease).49 The association may reflect shared etiopathogenetic mechanisms.
EoE with functional gastrointestinal disease	Concomitant functional chest pain, functional heartburn, functional dysphagia, and nonulcer dyspepsia may confound symptom-based treatment approaches for EoE.
EoE with tracheoesophageal fistula	Esophageal eosinophilia present in up to 18% of patients with history of tracheoesophageal fistula; these patients were reported to share an EoE-risk genotype.52

EGID, eosinophilic gastrointestinal disease; GERD, gastroesophageal reflux disease; IBD, inflammatory bowel disease.

Future Directions

What will care of patients with EoE look like in 10 years? We anticipate that predictive, preventative, precision, and personalized medicine will improve patient care. We might be able to integrate medical and dietary treatments with targeted therapies and office-based disease activity monitoring.⁵³ Multidisciplinary, multicenter, basic, translational, and clinical research, using a variety of model systems (in vitro, ex vivo, organoids, and animal models) combined with prospective clinical studies, will continue to transform care. Populations at risk of developing eosinophilic gastrointestinal diseases might be identified based on evaluation of family history, birth history, the genome, the microbiome, the exposome, pharmacogenetic features, disease course, and biomarker profiles.⁹ Studies have identified associations between early-life risk factors and EoE (such as antibiotic use in infancy, cesarean section, and preterm birth) and genetic variants and EoE (such as polymorphisms in TSLP, LOC283710, KLF13, CAPN14, CCL26, and TGFB).⁵⁴ Analyses of these risk factors might be used to predict disease course and select treatment. Treatments might include use of probiotics, specific diets, and targeted therapies. Periodic monitoring of disease activity and treatment side effects, along with patient education, are required to prevent EoE development and progression.

We need to learn more about the specific phenotypes and endotypes of EoE. Researchers have correlated clinical and pathology features with gene expression profiles to classify EoE subtypes.⁵⁵ Continued analyses of disease subtypes and biomarkers of inflammatory and remodeling processes might lead to identification of targets for therapy. Measurements of eosinophil granule proteins, bromotyrosine, cytokines, and IgG4 in blood, urine, or esophageal luminal contents are being evaluated to identify biomarkers and therapeutic targets. Artificial intelligence and other technologies might be used in diagnosis and selection of therapy, but the personal aspects of clinical care are needed to remove barriers to care and promote patient well-being.

Treatment plans for patients should be selected based not only on EoE features, but also the ability of patients, caretakers, and providers to implement the therapy. Education should address the aspects of eosinophilic gastrointestinal diseases specific to each patient and be delivered in a format best suited to patients’ learning styles. Less-invasive, office-based tools for assessment could replace endoscopy and biopsies. Cell and organoid systems might be used to identify specific environmental factors and foods that induce EoE in individual patients and to select treatments. Continuous, real-time monitoring with biosensors might be used to evaluate adherence to, and efficacy of, treatment regimens, thereby allowing for adjustments before relapse.

During the course of the last 30 years, EoE has progressed from a clinical curiosity to a well-recognized disease with many challenges to treatment. Guidelines for diagnosis have changed rapidly, and prospective studies of new treatment approaches are underway. Most patients are still treated with topical steroids or restrictive diets. However, studies have reported the efficacy of biologic agents against cytokines that contribute to EoE, and further studies will be necessary to determine how and when these can be used in treatment. Further studies into the pathogenesis of EoE could lead to agents that can be used to treat patients based on their specific features of this heterogenous disease.

Abbreviations used in this paper:

AGA-JTF

American Gastroenterological Association Joint Task Force on Allergy and Immunology

EoE

eosinophilic esophagitis

eos/hpf

eosinophils per high-power field

FDA

Food and Drug Administration

IL

interleukin

PPI

proton pump inhibitor

SFED

six-food elimination diet