Eosinophilic Esophagitis and the Eosinophilic Gastrointestinal Diseases: Approach to diagnosis and management

Abstract

The eosinophilic gastrointestinal diseases (EGID) represent disorders of the GI tract that result from the local infiltration and aberrant activity of eosinophils and other immune cells. Eosinophilic esophagitis is the most well-characterized EGID and is defined by the presence of intraepithelial eosinophils in the esophagus (≥15 eosinophils per high powered field) and clinical symptoms associated with esophageal dysfunction. The other EGID are rare and lack strong data regarding pathogenesis and management. The incidence and prevalence of EoE are increasing, and EoE is now a major cause of upper GI morbidity. Management is multidisciplinary, with collaboration between gastroenterologists, allergists, pathologists, and dieticians, and is aimed at amelioration of symptoms and prevention of long-term complications such as esophageal stricture. Treatment options for EoE include proton pump inhibitors, swallowed topical corticosteroids, and elimination diets. Esophageal dilation is used when esophageal strictures or fibrostenotic changes are present. Additional therapies targeting eosinophils and other mediators of Th2 inflammation are under development and are promising. Treatment options for other EGIDs typically involve corticosteroids or dietary elimination.

Introduction

Eosinophilic gastrointestinal diseases (EGID), including eosinophilic esophagitis (EoE), gastritis (EG), gastroenteritis (EGE), and colitis (EC), are characterized by GI tract eosinophilia and symptoms that cause significant morbidity in children and adults. EGID definitions and management are evolving as more is learned about the etiology and natural history of these disorders. EGIDs are chronic disorders that are thought to develop in response to an immunogenic trigger. The diagnosis of EGID is contingent upon exclusion of other disorders associated with eosinophilia. This review focuses on the current clinical guidelines, controversial topics, and emerging therapies for the best characterized and most common EGID, EoE. In addition, a brief discussion of what is known, and unknown, about EG, EGE, and EC is presented.

What is known about EGID: EoE

EoE Diagnosis

EoE was recognized as a distinct clinical entity in the early 1990s (1, 2), and since then incidence and prevalence have markedly increased (3). Initial diagnosis and management guidelines were released in 2007 (4) and updated in 2011 (5), 2013 (6), and 2017 (7). Early definitions of EoE required symptoms of esophageal dysfunction and esophageal eosinophilia on biopsy (at least 15 eosinophils per high-power field, eos/hpf), not otherwise explained by a potential competing cause of eosinophilia, be present after a high-dose proton pump inhibitor (PPI) trial. Significant research advances over the past five years, especially those related to the understanding of the role of PPIs (8), led a European task force to eliminate the requirement for a PPI trial. Therefore, those with esophageal eosinophilia who respond to PPI therapy (PPI-responsive esophageal eosinophilia, PPI-REE) exist on the continuum of EoE. This decision was re-emphasized during a recent international consensus conference (9) based on substantial evidence of overlap between PPI-REE and EoE in terms of clinical symptoms, endoscopic findings, pathophysiology, and molecular profiles; updated and operationalized diagnostic guidelines will be released this year.

EoE Pathophysiology

EoE is a non-IgE-mediated allergic immune response. It occurs more often in male patients (3:1 male:female), and its highest prevalence is currently in the 3^rd – 5^th decades of life (3). Clinical symptoms differ in young children compared to older children and adults, partially due to the progression of EoE from an inflammatory to fibrostenotic phenotype over time (10–12). Young children have feeding difficulties, reflux-like symptoms, vomiting, abdominal pain, food refusal, and failure to thrive (13, 14); older children and adults experience dysphagia, heartburn, chest discomfort, exercise-induced chest pain, and food impaction (15–20).

The natural history of EoE appears to progress from an inflammatory to fibrostenotic phenotype (11, 12, 21–27), but subepithelial fibrosis is detected even in children, suggesting that esophageal remodeling occurs early in the disease process (11, 28). Esophageal remodeling may also contribute to esophageal dysmotility in EoE (27, 29–36). Patients with EoE demonstrate impaired esophageal epithelial barrier integrity (27, 37–41) and increased esophageal sensitivity to acid (42) and local allergen exposure (43). Treatment of esophageal inflammation, either with topical corticosteroids or elimination diets, likely prevents long-term fibrostenotic changes and improves impaired barrier integrity in patients with EoE (39, 41, 44–53). Evidence does not currently support EoE as a premalignant lesion (54), although there are small case studies suggesting an association between esophageal eosinophilia and granular cell tumors (55–59).

The prevalence of atopic disease such as allergic rhinitis, bronchial asthma, IgE-mediated food allergies, and eczema, is far higher in patients with EoE than the general population (60), suggesting a prominent role for the Allergist in treatment of these patients’ allergic comorbidities (61). However, the definition of “food allergy” varies widely across studies (60, 62), which highlights a fundamental misunderstanding of the nature and role of skin prick testing (SPT) and serum food-specific IgE testing in diagnosing IgE-mediated food allergies. Allergists are uniquely qualified to determine “probable”, “possible”, and “unlikely” culprit food allergens from the clinical history and epidemiology of food allergies, which then informs subsequent testing. The 2011 guidelines recommended patients with EoE who previously demonstrated sensitization to a particular food based on allergy testing undergo office-based oral challenge prior to reintroduction of that food into their diet (5), as there are case reports of patients with EoE who develop an IgE-mediated food allergy after avoiding their EoE trigger food (63–67). On the other hand, there are case reports of patients with IgE-mediated food allergy on oral food immunotherapy who develop EoE, although the presence of EoE prior to starting this immunotherapy is not known (68, 69). A recent literature search estimated the prevalence of EoE in patients during oral immunotherapy for IgE-mediated food allergy to be approximately 5% (70). EoE itself is understood as a non-IgE-mediated food allergy; therefore, allergy testing to guide treatment is controversial and discussed below under the Elimination Diet section.

There is evidence suggesting that exposure to aeroallergens may contribute to the pathogenesis of EoE in some individuals, although it is unclear whether aeroallergens alone can cause EoE or if exposure can modify disease in certain patients with food-triggered EoE (71). In a retrospective chart review Gita Ram, et al., identified a subset of patients with EoE and aeroallergen sensitization whose EoE symptoms and histopathologic findings worsened during the season corresponding to their specific aeroallergen sensitization in the absence of dietary or treatment changes (72, 73). Interestingly, there are case reports of patients on oral or sublingual aeroallergen immunotherapy who develop EoE that subsequently resolves with the cessation of the immunotherapy (74–76). It is not known whether EoE disease activity improves in these patients with more aggressive treatment of allergic rhinitis and with counseling on allergen avoidance, but one intriguing case report showed resolution of EoE in one patient after two years of dust mite oral immunotherapy (77). Therefore, the Allergist’s role in management of EoE is multifaceted.

EoE Monitoring

Several validated scoring systems measure symptoms and disease activity in EoE, though most are currently being used primarily for research purposes (78–81). Scoring systems focused on tracking changes in endoscopy/histology findings are being increasingly used in clinical practice, including the EoE Endoscopic Reference Score (EREFS; Figure 1) (82, 83). This classification is increasing being used in endoscopic reports and represents an important way to monitor endoscopic severity over time.

Figure 1.

A range of endsocopic findings in EoE with application of the EREFS classification, which measures the 5 main endoscopic features of EoE: Edema, Rings, Exudates, Furrows, and Strictures. Edema is graded as absent (0) or present (1); Rings are graded as absent (0), mild (1), moderate (2), or severe (3); Exudates are graded as absent (0), mild (1), or severe (2); Furrows are graded as absent (0), mild (1), or severe (2); Stricture are graded as absent (0) or present (1), and if present the inner diameter can also be reported. (A) A patient with suspected EoE, but with a normal endoscopy. (B) A patient with edema, exudates, and furrows. (C) A patient with edema, rings, exudates, furrows, and a stricture. (D) A patient with edema, with edema, rings, exudates, furrows, and a stricture. (E) A patient with edema and furrows. (F) A patient with edema, rings, exudates, and furrows.

There can be a disconnect between EoE symptoms and endoscopic or histologic measures of disease activity. For example, patients may be able to minimize symptoms with dietary avoidance or modification (careful chewing, slow eating, avoiding hard or fibrous foods) despite ongoing inflammation, or conversely, if there is an esophageal stricture, symptoms may persist despite resolved inflammation. Therefore, close clinical follow-up of patients is required, and histopathology remains necessary for monitoring EoE disease activity. However, studies have used varying thresholds of eos/hpf to determine treatment response (84). Recent work supports histologic response as an eosinophil count of <15 eos/hpf, as it is modestly predictive of gross endoscopic or symptomatic improvement, but a more stringent level of <5 eos/hpf correlates with combined improvement in both of these parameters (85, 86). While EoE is a chronic disease, the optimal intervals for endoscopic surveillance of patients who have achieved remission is not known and should be an area of future study. However, endoscopic assessments with esophageal biopsy should be made approximately 6–8 weeks after a treatment is initiated or changed in order to evaluate treatment efficacy as measured by endoscopy and histologic response. Once the disease is under control, less frequent intervals are acceptable. There are no guidelines as to the recommended frequency, so surveillance endoscopy may be performed at intervals that is dictated by the clinical picture. Some clinicians opt to do this on a yearly basis, but others perform endoscopy less frequently (87).

EoE Treatment

Management of EoE requires coordination between providers across disciplines and can be time- and resource-intensive. Ultimately, a patient-centric discussion will determine the best approach for individualized care, and gastroenterologists, allergists, dieticians, and nurses are integral to this process. While there are no FDA-approved treatments for EoE, current options include PPI therapy, swallowed topical corticosteroids (tCS), and elimination diets (88, 89). Esophageal dilation is an additional option for treatment of esophageal strictures and fibrostenotic changes.

Proton Pump Inhibitor Therapy

PPIs are inexpensive, effective, and despite recent reports of adverse effects, are still considered generally safe for long-term use (90–92). A novel anti-inflammatory/anti-eosinophil mechanism of PPIs is independent of acid suppression; indeed, histologic remission occurs in up to half of patients with symptomatic esophageal eosinophilia on PPI therapy (93–99), only a proportion of whom have gastroesophageal reflux. Interestingly, the presence of acid reflux does not predict who with EoE will respond to PPI treatment. Future studies may find ways to identify which patients with EoE will respond to PPI therapy as a first-line treatment. Omeprazole 20–40 mg twice daily or its equivalent in adults, or 1 mg/kg/day (max 40 mg once daily) is considered high dose PPI therapy in the setting of esophageal eosinophilia (6), and in a meta-analysis induced histologic remission in up to 50% and symptomatic improvement in up to 60% of adults and children (93, 100, 101). Twice daily dosing is more effective at inducing histologic response than once daily dosing (100). Sustained remission is achieved in more than three-quarters of patients, and a majority of those who relapse on maintenance dosing can achieve remission at a higher treatment dose (101–103). Thus, PPI therapy can be considered as an initial treatment. An important caveat is that most of the known treatment responses of steroids and dietary elimination are in patient populations who were non-responsive to PPI use, and the comparative efficacy of PPIs, steroids, and diet is not known (104). In Table 1, we provide dosing recommendations for PPI therapy in children and adults based on our experience, though we acknowledge there are currently no FDA-approved dosage guidelines for PPI therapy in EoE.

Table 1. UNC Experience: Recommendations for tCS and PPI therapy.

Topical Corticosteroids (tCS)  Dosing typically divided twice daily
	Initial Dosing  6–12 week period		Maintenance Dosing*		Notes
Drug	Children	Adolescents &  Adults**	Children	Adolescents  & Adults**
Fluticasone  Proprionate 220  mcg Inhaler	880–1760  mcg/day	1760 mcg/day	440–880  mcg/day	880–1760  mcg/day	No eating/drinking/brushing teeth for 30 minutes after each dose Do not use a spacer: Puff directly into mouth & swallow
Budesonide  Slurry  0.25 mg/2 mL  respules  OR  0.5 mg/2 mL  respules; PLUS  thickening  agent***	1 mg/day	2 mg/day	0.5  mg/day	1 mg/day	No eating/drinking/brushing teeth for 30 minutes after each dose Use 0.25 mg respules instead of 0.5 mg respules if a larger volume desired
Proton Pump Inhibitors (PPI)
Drug	Children <10 years old		Adolescents & Adults		Notes
Lansoprazole	2 mg/kg/day max divided bid		30 mg twice daily		Initial treatment 8–12 weeks to  assess response  If response, can attempt to  wean dosing and evaluate for  minimal effective clinical dose  Ideally give PPI 30 minutes  before meals
Omeprazole	2 mg/kg/day divided bid		20–40 mg/day divided bid
Pantoprazole	2 mg/kg/day max divided bid		20–40 mg twice daily
Esomeprazole	2 mg/kg/day max divided bid		20–40 mg once daily
Rabeprazole	--		20 mg twice daily
Dexlansoprazole	--		60 mg daily

^*Goal is to use the lowest effective dose, but few data to support this approach or any maintenance approach; recent literature suggests that some patients relapse with tCS dose reduction (118)

^**Some have used cut-off from children to adults as >5 feet tall (109)

^***Example regimens include mixing contents of each respule with 5 grams of sucralose (1 packet of sucralose = 1 gram) or 2.5 mL Neocate Nutra per mg of budesonide to make total volume of 8–12 mL (96)

Swallowed Topical Corticosteroids (tCS)

Multiple systematic reviews and meta-analyses show that tCS induce remission in EoE and are safe and well-tolerated in the short term, although studies vary in medication type, dosing, method of administration, and duration of treatment (105–109), and discontinuation of therapy results in recurrence of active disease (87). Currently available tCS adapt the use of existing asthma preparations for administration to the GI tract. Instead of inhaling fluticasone propionate (FP) for asthma, use for EoE requires spraying the MDI directly to the mouth followed by swallowing. Major placebo-controlled prospective studies examining FP in children used daily doses between 880–1760 mcg per day, divided twice daily (110, 111). Low doses of FP in children may also be supported; partial or complete histologic remission was achieved in 63–83% of children using low swallowed FP dosing, based on age-specific treatment doses for asthma (2–4 years old: 176 mcg/day, 5–11 years old: 440 mcg/day, and ≥12 years old: 880 mcg/day) (112). In adults, high dose FP therapy (1760 mcg/day) when compared to placebo therapy achieved a 90% decrease in mean baseline esophageal eosinophilia in 62% of patients (113). Low dose FP therapy (880 mcg/day) was less effective at inducing histologic remission (114, 115) compared to PPI, though this was not directly compared to high-dose FP therapy.

To minimize pulmonary deposition and maximize esophageal deposition of tCS, budesonide respules can be mixed with either sucralose (5 grams per 2 mL respule), elemental formula, or other thickening agents (116), to prepare a slurry for patients to swallow. The first prospective study examining budesonide slurry (children shorter than 5 feet received 0.5 mg/day and children taller than 5 feet received 1 mg/day) showed reductions in esophageal eosinophilia in 87% of patients (117). Among adult populations, nebulized budesonide 1 mg twice daily for 15 days resulted in histologic remission in 72% of patients with EoE (118). However, oral viscous budesonide 1 mg twice daily was more effective than nebulization of the same dose, corresponding with increased esophageal mucosal contact time as measured by scintigraphy (119). This finding led to the development of novel methods of delivering tCS directly to the esophageal mucosa specifically for EoE. Budesonide effervescent tablets (BET) were effective in inducing histologic remission in 100% and 94% of patients receiving 1 mg BET twice daily or 2 mg BET twice daily, respectively (120), and has recently been approved for EoE treatment in Europe. Pre-mixed viscous budesonide, currently in phase 3 trials, induced histologic remission of EoE and reduced clinical symptoms in both children and adults (121, 122).

After initial treatment and response to tCS, there are relatively few data to guide long-term treatment (87). One strategy is to decrease the dose to the lowest effective level to minimize medication side effects. In one study, high dose FP (1760 mcg/day: 4 puffs of the 220 mcg MDI twice daily) for 12 weeks induced histologic remission in 65%–77% of patients with EoE. Among those who responded to FP therapy, a 50% dose reduction (4 puffs of the 220 mcg MDI once daily) was effective at maintaining histologic remission in 73%–93% of patients (110). This study also demonstrated that patients who failed 12-week high dose FP treatment were unlikely to respond to a prolonged course of this same treatment. Intermittent dosing is not supported by a small study of budesonide slurry every Monday-Wednesday-Friday, which found the majority of patients relapsed (123). A randomized, double-blind, 50-week trial of 0.5 mg daily budesonide maintenance therapy versus placebo showed 36% of patients in the active arm compared to 0% of patients on placebo who maintained remission (124). A retrospective cohort study (125) was performed on patients who were able to achieve “deep remission”, defined as combined clinical, endoscopic, and histologic remission, for six months on tCS. Of the 33 patients (9.4%) who achieved deep remission, only six (18.2%) maintained deep remission off tCS. Other studies have raised the question of whether there might be a loss of response or steroid resistance in some patients during an extended treatment course (52, 126, 127).

In general, tCS appear to be safe and well tolerated, but there are potential side effects. A recent meta-analysis found a 4–5% rate of esophageal candidiasis among children and 5–15% in adults (106). To date, no study has described decreased growth in pediatric patients with EoE receiving tCS (112, 121, 128). While there are some reports of adrenal suppression due to corticosteroid use in EoE (129), a systematic review found that clinically important adrenal insufficiency is rare (130). Given the association of EoE with other atopic diseases that predispose to glucocorticoid use, cortisol monitoring to detect adrenal insufficiency may be considered for children with EoE if they are receiving long-term high-dose tCS therapy or are receiving different steroid formulations for other atopic conditions, but there are no specific recommendations in guidelines to date. In Table 1, we provide dosing recommendations for tCS therapy in children and adults based on our experience.

Elimination Diets

The first clue that EoE was a food antigen-driven allergic disease came from a landmark study where 10 children with esophageal eosinophilia who did not respond to PPI treatment were treated with an elemental diet; 8 had complete resolution of symptoms, and 10 had significant reduction in eosinophil counts (131). Reintroduction of foods in a controlled setting identified trigger foods, and recurrence of GI symptoms occurred. An elemental diet is amino acid-based and devoid of intact food proteins. While the elemental diet is effective (a meta-analysis found 91% histologic response (132)), there remain significant barriers to widespread implementation of this diet, including poor palatability, high financial cost, inconsistent health insurance coverage, and increased psychosocial isolation of the patient (81, 133–136). Thus, alternative elimination diets have been developed that balance time to remission and to identification of the food trigger, minimize number of repeat endoscopy with biopsies, and improve patient adherence.

Serum food-specific IgE, skin prick testing (SPT), atopy patch testing (APT), or combinations thereof, have poor positive predictive value of non-IgE-mediated food triggers in adults with EoE (137, 138). Allergy test-directed elimination diets perform poorly in EoE, inducing histologic remission in approximately 43% of patients with EoE with high variability between studies suggesting low reproducibility (132, 138). However, in the context of a detailed history of possible food triggers along with epidemiologically known culprits, history-directed allergy testing, especially when APT is done by rigorously trained individuals, may help identify food triggers in some patients with EoE (139–141). Practically speaking, APT is technically challenging, and empiric elimination diets have been shown to be effective for the treatment of EoE. In 2006, a retrospective study (142) compared an elemental diet with a diet that eliminated the six most common allergenic foods (the so-called six food elimination diet (SFED)): cow/animal-milk protein, soy, egg, wheat, peanuts/tree nuts, and seafood (143). After six weeks, 74% and 88% of children on the SFED and elemental diet, respectively, achieved histologic remission (<10 peak eos/hpf). This approach was used in other studies, and a meta-analysis showed the SFED is successful at inducing histologic remission in 72% of patients with EoE (132). However, this diet is quite restrictive, requires numerous endoscopies to monitor disease activity and to reintroduce individual foods, and there remains controversy about the extent of elimination with certain food groups (for example, soy or all legumes, or wheat or all cereal grains) (144).

The most common EoE triggers have been reported to be animal-milk protein (adults: 39–64%, children: 65–85%), wheat/gluten-containing cereals (adults: 22–60%, children: 26–37%), egg (adults: 5–44%, children: 17–40%), and soy/legumes (adults: 7–24%, children: 10–38%) (145); 65–85% of patients had a single food trigger (146). The four-food elimination diet (FFED, animal milk, wheat/gluten, egg, soy/legumes) and two-food elimination diet (TFED, animal milk, wheat/gluten) achieved histologic remission in 54% and 43% of adult patients, respectively, and in 64% and 43% of pediatric patients, respectively (132). In a multicenter prospective study (147) utilizing a step-up approach, clinicohistologic remission was achieved in 43%, 60%, and 79% of patients on the TFED, FFED, and SFED, respectively. This approach also had the potential to decrease the number of endoscopies by 20% and time to identification of food trigger(s) by 30% when compared to a SFED. Importantly, a detailed, user-friendly guide was developed for this study (147) that can serve as an important clinical tool as empiric elimination diets become more widely used. Table 2 provides a summary of the TFED, FFED, and SFED.

Table 2. A Practical Guide to the Empiric Elimination Diet in the US^a.

Diet	Two-food group	Four-food group		Six-food group
Eliminated  food groups	Animal milk, wheatb	Animal milk, wheatb, egg,  soyb		Animal milk, wheatb, egg,  soyb, nuts, fish/seafood
Allowed  food groups  (favor raw,  fresh, or  uncooked)	Vegetables, tubers, non-  heat grainsc, legumesd,  meate, egg, fish/seafoodf,  fruit, nuts, [dairy made from  soy, rice, spelt, quinoa,  walnut, nuts]	Vegetables, tubers, non-  wheat grainsc, non-soy  legumesd, meate,  fish/seafoodf, fruit, nuts, [dairy  made from rice, spelt quinoa,  walnut, nuts]		Vegetables, tubers, non-  wheat grainsc, non-soy  legumesd, meate, fruit, [dairy  made from rice, spelt, quinoa]
	Foods include:		Food labels may contain:
Animal  milk	Cow’s, goat’s, and sheep’s milk and milk  products (cheese, yogurt, butter,  margarine, ice creams, milkshakes,  custard, crème caramel, rice pudding)		Milk, cream, caseinates, hydrolysates, lactalbumin,  casein, whey, E-4511, E-4512, E-4513, flavor,  animal fat, cream, proteins, dehydrated powder or  sauce
Wheat	Wheat, wheat-containing: bread, toast,  biscuits, cookies, donuts, muffins, pretzel,  pancakes, waffles, crackers, cream  desserts, sweets, candies, pasta, cream,  soups, sauces, malted food, breaded or  floured vegetables, [beer, whiskey]		Flour or floured, farina, wheat-enriched, malted or  malt added, breaded, E-1404, E-1412, E-1413, E-  1414, E-1420, E-1422, E-1440, E-1442, E-1450,  starch, fiber, protein, vegetable protein, hydrolyzed  protein, malt, malt extract, cous cous, yeast,  species, flavour
Egg	Egg, baked goods, pasta, cakes, biscuits,  cookies, donuts, muffins, pretzel,  pancakes, waffles, crackers, cream  desserts, sweets, candies, processed  meat, goose liver, mayonnaise, coat and  wrapped in bread food, breaded or  creamed vegetables, sauces		Albumin, apovitellin, binder, coagulant, cholesterol  free egg substitute, dried egg, egg white, egg yolk,  egg lecithin, egg lysosome, eggnog, egg wash,  globulin, lecithin, livetin, lysozyme, meringue,  meringue powder, simplesse, surimi, ovoalbumin,  ovamucin, ovamucoid, ovotransferrin, ovovitelin,  powdered egg, trailblazer, vitellin, whole egg
Soyg	Soy sauce, tofu, edamame, tempeh,  vegetarian meat analogs		Oil made with soy, hydrolyzed plant, vegetable  protein, plant protein, vegetable gum, vegetable  starch
Nutsg	Almond, artificial nuts, Brazil nut,  beechnut, butternut, cashew, chestnut,  chinquapin nut, coconut, filbert/hazelnut,  macadamia nut, marzipan/almond paste,  Nangai nut, natural nut extract, nut butters,  nut meal, nut meat, nut milk, nut paste, nut  pieces, pecan, pesto, pili nut, pine nut  (e.g., Indian, pignoli, pinolia, pignon, pinon,  and pinyon nut), pistachio, praline, shea  nut, walnut		Oil made with any of the named nuts to the left, nut  flavoring
Fish/ Seafoodg	Fish (anchovies, bass, catfish, cod,  flounder, grouper, haddock, hake, halibut,  herring, mahi mahi, perch, pike, pollock,  salmon, scrod, swordfish, sole, snapper,  tilapia, trout, tuna), shellfish (crab, lobster,  prawns, shrimps), and mollusks (cockles,  mussels, octopus, oyster, snails, squid)		Oil or gelatin made with named seafood to the left

^aAdapted from Molina-Infante, J, et al., 2017{Molina-Infante, 2017 #3}, and Kliewer, KL, et al., 2017 {Kliewer, 2017 #159}, for the US population.

^bExpanded elimination diets, which exclude wheat/gluten/cereal grains and all legumes in place of wheat and soy, respectively, have been proposed despite no studies exploring whether these increasingly restrictive elimination diets are more effective or clinically relevant for EoE treatment {Kliewer, 2016 #184}. In the case of grains, there is a risk of: 1) cross-reactivity between similar grain proteins, and 2) cross-contamination due to processing of different grains within the same facility. However, the clinical relevance of these concerns is not known. Diverse legumes are consumed by Spanish populations and are frequently associated with EoE in these patients {Molina-Infante, 2018 #218}. The provider may take this into consideration when advising a patient to avoid soy or all legumes.

^cNon-wheat grains include barley, rye, gluten-free oats (due to high risk of cross-contamination during the processing of conventional oats {Kliewer, 2017 #159}), spelt, triticale, semola, semolina, and kamut.

^dLegumes include soy, lentils, pea, chickpeas, beans, peanuts, lupin, guar gum, carob bean, and alfalfa.

^eExcepting processed or pre-cooked meats, such as sausages or hamburgers.

^fExcepting processed or pre-cooked fish.

^gAfrican and Asian ethnic foods often contain legumes (soy), nuts, and fish/seafood, and may be high-risk foods.

The reintroduction of foods after achieving histologic remission has not been studied in detail and varies among providers. Sequential addition of individual foods into the diet for a period of about six weeks should be followed by endoscopy with biopsy to monitor disease activity. A common method is to introduce two or more foods sequentially at two-week intervals, depending on how many foods were initially eliminated, before undergoing endoscopy at regular intervals. For example, a patient on a SFED may first introduce seafood for two weeks; if asymptomatic, the patient can introduce peanuts/tree nuts for two weeks, followed by an endoscopy to assess the histologic response to these two food groups. For patients who had more extensive elimination diets, introduction of less allergenic foods (lower protein content such as fruits and vegetables) are typically tried before more allergenic foods. An example of a food reintroduction schedule is shown in Figure 2 (104). Foods from each category can be reintroduced every five to seven days (148). The patient should ingest an age-appropriate form and serving size amount of the food on most of the days of the week during the trial period (145). If a patient develops recurrence of symptoms with food reintroduction, they may undergo a wash-out period of six weeks prior to reintroduction of the next food group (149). Of note, if two foods are added back and recurrent esophageal eosinophilia is noted, then it is not possible to determine which of the foods is a trigger. Because of that, the more highly allergenic foods (i.e. Column D of Figure 2) are typically added back individually followed by repeat endoscopy in 6 weeks to confirm histologic remission prior to further food introductions.

Figure 2.

Dietary reintroduction of food allergen groups for patients with EoE. Modified with permission from Gastrointest Endosc Clin N Am. 2008;18(1):179–94; xi (148) and Gastroenterology. 2014;147(6):1238–54 (96).

Complete avoidance of the culprit food should maintain long-term symptomatic and histologic remission without the need for pharmacologic therapy (6, 7). There remains, however, a dearth of studies addressing the efficacy and feasibility of long-term food allergen avoidance in EoE. Adult patients who avoided their food trigger(s) were able to achieve sustained symptomatic and histologic remission for up to three years without the need for pharmacologic therapy (47, 150). However, long-term compliance can be difficult, and lower long-term “real-world” response rates have also been noted (151). In addition, patients with two or more food triggers more often preferred pharmacologic therapy than a continued restricted diet (88).

There are several factors that should be considered before embarking on an elimination diet, particularly in pediatric patients. The choice to pursue dietary therapy may be influenced by the severity of disease, nutritional status of the patient, learned maladaptive feeding behaviors, family dynamics, financial resources, and family preferences (145). Learned maladaptive feeding behaviors are highly prevalent in pediatric patients with EoE and can persist despite histologic remission (152). Many children also exhibited delayed oral motor skills required for feeding, as EoE during early formative years hinders the natural development of these skills (153). Furthermore, compensatory behaviors such as thoroughly chewing food prolongs meal times and may limit caloric intake (154). Individual or group therapy with a pediatric feeding specialist provides re-learning of appropriate feeding behaviors, development of oral motor and sensory skills, allows interaction with peers, and provides communal support for caregivers. Dieticians are key team members for EoE treatment who provide practical education, guidance to ensure adequate nutritional intake, knowledge of community resources, and a unique perspective of the patient’s barriers to care. Current recommendations encourage the involvement of a dietician and other support services if warranted (6, 7).

Esophageal Dilation

It is estimated that esophageal strictures are present in 30–80% of adults with EoE, and prevalence increases proportionally with age and delay in time to diagnosis (25). Multiple meta-analyses and systematic reviews support the safety and efficacy of endoscopic dilation in adult and pediatric patients with EoE (155–157). Some patients, especially those with smaller initial esophageal diameters, will require multiple procedures over several months to achieve improvement in esophageal caliber; others will need repeat dilation due to stricture recurrence (158). Importantly, endoscopic dilatation does not alter the underlying inflammatory process in EoE, and treatment with concomitant pharmacologic therapy or elimination diet reduces the risk of needing subsequent dilations (26, 59). There are no definitive guidelines for when to offer endoscopic dilation, although it is generally recommended for patients with EoE who have food impaction, strictures, or whose dysphagia does not respond to therapy, or when critical strictures are present, even at the time of diagnosis.

Challenges that remain in EGID

Refractory EoE

Refractory disease in EoE is conceptually defined as ongoing clinical symptoms, endoscopic findings, and esophageal eosinophilia despite treatment with either a tCS or elimination diet (159–163). A significant proportion of patients in both observational studies and randomized controlled trials of EoE meet these criteria (159). There is no clear consensus on specific endpoints or response thresholds that should define treatment failure. A recent review on this topic proposed that refractory disease constitutes incompletely resolved esophageal eosinophilia (≥15 eos/hpf) in the setting of persistent endoscopic findings and clinical symptoms (159). However, patients may have discordant findings among these domains, and it is not clear how these patients should be categorized. For instance, a patient may have complete resolution of histologic and endoscopic findings but continue to experience dysphagia. Importantly, the patient’s individual pattern may yield clues as to the underlying cause for their refractory disease; in the previous example of a patient with continued clinical symptoms, one should consider the presence of esophageal strictures.

The first step in managing patients with refractory EoE is to consider patient nonadherence, suboptimal medication dose, ineffective delivery or inappropriate administration of the medication, continued exposure to allergen, concomitant infection, esophageal stricture, or an inappropriate diagnosis of EoE. Patient nonadherence is an especially important consideration in elimination diets, as this therapy is restrictive, expensive, and requires significant nutritional literacy.

Once a patient has confirmed refractory EoE, a careful consideration of the patient’s initial failed treatment (tCS versus elimination diet), disease severity, clinical course, and individual biopsychosocial circumstances should guide treatment. Unfortunately, there are few evidence-based data to guide the management of refractory EoE. Therefore, recommendations are based on logical and systematic progression of therapies. For patients who failed tCS, the dose, mode of delivery, and specific corticosteroid may need adjustment. The addition of an elimination diet or second-line therapy in combination with tCS may be a reasonable approach. Second-line treatments could include systemic corticosteroids, leukotriene antagonists, mast cell stabilizers, immunomodulators, or biologic agents. Patients refractory to an elimination diet should consider the elemental diet, despite significant downsides to this therapy, as it is highly effective (132). Many patients with treatment non-response may be good candidates for clinical trials. A diagnostic and management algorithm for patients with EoE is provided in Figure 3.

Figure 3.

Algorithm for diagnosis and management of EoE. “Response” refers to <15 eos/hpf with resolution of symptoms. “No Response” refers to >15 eos/hpf and persistence of symptoms. For patients with a histologic response but with ongoing symptoms, assessment should be made for a persistent esopahgeal stricture or alternative cause of symptoms (dysmotility, viserceral hypersensitivity, etc). Patients with no histologic reponse but with resolution of symptoms may have undergone prior esophageal dilation and should be assessed for dietary avoidance or modification behaviors.

EGIDs: EG, EGE, and EC

In comparison to EoE, data related to the non-EoE EGIDs are immature, with most information derived from single center case series; there are no RCTs or published guidelines.EG, EGE, and EC are rare, with an estimated prevalence of ~5–10 per 100,000, accounting for ~50,000 cases in the U.S (164, 165). While there is an association with atopic diseases as is seen in EoE, patients with EGID have a unique gene expression signature that does not fully overlap with EoE (166). In addition, there does not seem to be the same male predominance as in EoE.

In general, the presentation of EG, EGE, and EC is dependent on the location of the eosinophilic infiltration (167–172). Symptoms of EG include abdominal pain, nausea, and vomiting. Symptoms are similar for EGE, though diarrhea can also be seen. For EC, symptoms include abdominal pain, diarrhea, and hematochezia. Depending on the extent of eosinophil infiltration though the GI tract wall, there can be malabsorption, protein-losing enteropathy, or ascites. The diagnosis is established after endoscopic biopsies show increased levels of eosinophils in the GI tract and excluding other potential causes of eosinophilia. There is debate as to the eosinophil threshold for diagnosis, and this level depends on the area of the GI tract examined (173). Eosinophils are normally present in the stomach, small bowel, and colon, which is in contrast to the esophagus where they are not normally seen (174). Therefore, a careful histopathologic examination is required to demonstrate pathogenic infiltration. Experts suggest at least 30 eos/5 hpf in the stomach designates EG, at least 50 eos/hpf in the duodenum establishes EGE, and even higher levels in the colon, particularly on the right side, suggest EC (173).

Treatment of EG, EGE, and EC is challenging, particularly as data are limited to case series (170–172, 175). Systemic steroids are a mainstay of therapy, but side effects limit long-term use and disease typically recurs after steroid discontinuation. “Topical steroids” have also been used. For example, enteral release budesonide can be useful for eosinophilic enteritis and proximal colitis, and granules of budesonide can be crushed and swallowed for delivery to the stomach. However, the effect is variable. Dietary elimination has also been reported, with similar strategies as with EoE, including SFED and elemental formula (176). Finally, there are scattered case reports of treatment with immunomodulators and mast cell agents (cromolyn; ketotifen).

Emerging Therapies for EGIDs

There has been keen interest in developing new pharmacologic treatments for EoE and the EGIDs in recent years (177). In addition to the novel topical steroid formulations as noted above, the increasing knowledge of EoE pathogenesis has allowed a number of therapeutic targets to be identified and tested. The anti-IL-5 agents mepolizumab and reslizumab are currently approved for treatment of eosinophilic asthma and have been tested in phase 2 randomized controlled trials in EoE with a moderate impact on tissue eosinophilia (178–180). The anti-IL-5 receptor antagonist benralizumab is also approved for eosinophilic asthma and may be a future treatment for EGIDs. There have been two randomized controlled trials with anti-IL-13 agents (181, 182), the most recent of which was a phase 2 study with RPC4046 showing promising results. The anti-IL-4 receptor antagonist dupilumab, approved for the treatment of eczema, is also in a phase 2 trial of EoE with promising results (183). An anti-TSLP agent has recently shown efficacy for eosinophilic asthma and may be a future treatment for EGIDs. A small molecule antagonist of the chemoattractant receptor-homologous molecule on Th2 cells shows some beneficial effect in a small study (184). Notably, omalizumab and infliximab have not shown efficacy in EoE (185, 186).

Conclusions and Future Directions

Non-IgE mediated eosinophilic infiltration of the GI tract can cause significant morbidity in patients of all ages. This aberrant immune response may induce a progression from inflammatory to fibrostenotic disease in EoE. PPI therapy, tCS, and dietary elimination are effective at inducing remission in many patients with EoE, and adjunctive endoscopic dilation is helpful in cases with esophageal strictures or fibrostenotic changes. Treatment plans should be individualized based on patient values, financial resources, lifestyle, and social support. Future research should attempt to identify patients who will respond best to specific therapies. Novel therapies targeting eosinophil and Th2-mediated inflammation in EoE are currently in development and show therapeutic promise. The other EGIDs are rare and less well studied in terms of both pathogenesis and management. Better understanding of EGID pathogenesis will ultimately lead to the identification of clinically-meaningful outcomes that will inform the guidelines for diagnosis and management of these conditions.

Abbreviations:

APT

atopy patch test

BET

budesonide effervescent tablet

EC

eosinophilic colitis

EG

eosinophilic gastritis

EGE

eosinophilic gastroenteritis

EGID

eosinophilic gastrointestinal disease

EoE

eosinophilic esophagitis

EoEHSS

EOE-specific histologic scoring system

Eos

eosinophils

Eos/hpf

eosinophils per high-powered field

EREFS

EoE Endoscopic Reference Score

FFED

four food elimination diet

FP

fluticasone propionate

GERD

gastroesophageal reflux disease

GI

gastrointestinal

PPI-REE

proton pump inhibitor-responsive esophageal eosinophilia

SFED

six food elimination diet

SPT

skin prick test

tCS

topical corticosteroids

TFED

two food elimination diet