Food allergy and eosinophilic esophagitis: Oral immunotherapy reveals a disease in flux

Oral immunotherapy (OIT) is a proactive approach to modulate acute IgE-mediated hypersensitivity responses to food through incremental allergen exposure. Gastrointestinal adverse events (AEs) are common and can represent a major obstacle to successful desensitization. In a smaller proportion of patients, OIT may induce eosinophilic esophagitis (EoE), a chronic delayed hypersensitivity response associated with tissue eosinophilia and esophageal dysfunction. Notably, there have also been case reports of non-esophageal eosinophilic gastrointestinal diseases (EGIDs) developing during OIT. The de novo development of EoE during OIT provides a unique lens through which to examine EGID pathogenesis. Here we briefly discuss the epidemiology, clinical manifestations, and review potential management strategies. Further, we discuss the mechanisms underlying OIT-induced EoE and identify areas for future investigation.

The epidemiology of OIT-induced EoE is controversial and incidence estimates are complicated by detection bias, attrition bias, and heterogeneity. EoE is much more common in patients with atopy; therefore, comprehensive screening for gastrointestinal symptoms and esophagogastroduodenoscopy (EGD) is required to exclude prevalent cases. For example, a single-center, retrospective study reported an overall EoE prevalence of 2.8% in their OIT cohort (n=607). Of these, 35.3% reported symptoms of esophageal dysfunction before OIT¹. Moreover, most patients who develop gastrointestinal AEs during OIT never undergo upper endoscopy. Instead, OIT is discontinued, and gastrointestinal symptoms usually resolve². The prevalence of EoE after OIT has been estimated between 1-2.7%^{2, 3}. This is likely an underestimate when accounting for patients with gastrointestinal symptoms who never undergo EGD. Indeed, a meta-analysis of 2,424 OIT subjects, showed 5.6% of patients discontinue OIT due to EoE or symptoms possibly related to EoE⁴. We prospectively examined adult patients with IgE-mediated peanut allergy (n=21) with EGD prior to OIT and found 14% of patients had pre-existing esophageal eosinophilia [>15 eosinophils per high power field (eos/hpf)]. One subject developed tissue eosinophilia and a food impaction during OIT – demonstrating OIT can induce EoE⁵.

OIT-induced gastrointestinal AEs span a spectrum of clinical symptoms, likely reflecting underlying pathomechanisms. Patients may develop nausea, abdominal pain, and vomiting early in OIT. This can be immediate or delayed and preclude continued updosing. In contrast, patients may experience more classic symptoms of EoE with dysphagia and food impaction. Gastrointestinal AEs 2-8 hours after OIT dosing may be associated with peripheral blood eosinophilia [absolute eosinophil count (AEC) > 600] and have been termed OIT-induced gastrointestinal symptoms and eosinophilic responses (OITIGERs⁶) or EoE-like OIT-related syndrome (ELORS⁷). Risk factors for gastrointestinal AEs with peripheral eosinophilia include a higher starting dose of OIT, faster rate of escalation, and baseline peripheral eosinophilia⁶. Most patients undergoing OIT will be asymptomatic or have oropharyngeal itching. Interestingly, we observed transient esophageal eosinophilia in nearly all subjects undergoing OIT, mirroring a similar pattern in peripheral eosinophil counts. Both trends were observed irrespective of gastrointestinal symptoms^{5, 6}.

EoE diagnosed during OIT is usually reversible with discontinuation of OIT^{2, 3}. Some advocate managing gastrointestinal AEs during OIT with dose reduction, modification, or temporary pauses in lieu of discontinuation^{6, 7}. These adjustments may facilitate successful desensitization; however, some patients may have persistent EoE requiring indefinite therapy. Neither antihistamines nor omalizumab protect against gastrointestinal AEs. Proton pump inhibitors and steroids (swallowed topical and systemic) have been used successfully to treat OIT-induced EoE. Epstein-Rigbi et al. reported 92% of patients with OIT-related EoE continued OIT. The majority achieved complete EoE remission with standard-of-care pharmacologic therapies. The remainder demonstrated full clinical resolution, but only a partial histologic response⁸. Studies investigating the efficacy of dupilumab as an adjunct to OIT are ongoing (NCT04148352, NCT03793608).

A related scenario is whether to consider OIT in patients with pre-existing EoE. Hardwick et al.⁹ described six patients with known, but well-controlled, EoE who underwent OIT for comorbid IgE-mediated food allergy. Three patients tolerated OIT, though two required adjustments to EoE and/or OIT treatment regimens to maintain EoE remission. Of note, EoE is a chronic disease with involved treatment strategies and invasive diagnostic modalities. Untreated EoE may lead to progressive esophageal dysfunction, esophageal stricture, and food impactions. As new evidence emerges about the safety and feasibility of initiating or continuing OIT in the face of EoE, ethical principles should be applied and shared decision making should be employed¹⁰.

The commonality linking EoE and IgE-mediated food allergy is epithelial barrier dysfunction. In our cohort of patients with IgE-mediated peanut allergy, all patients had evidence of dilated intercellular spaces (DIS) at baseline. OIT induced or exacerbated DIS and basal cell hyperplasia in addition to tissue eosinophilia⁵. We hypothesize these features represent an epithelial stress response to local adaptive immunity stimulated by chronic antigen exposure (i.e., OIT). Recent evidence suggests that eosinophils are not the primary cell type driving EoE pathology; instead, antigen-specific pathogenic effector T helper type 2 (peTh2) cells promote local inflammatory responses predominantly via IL-13-mediated pathways. Importantly, IgE production in food allergy requires a different T cell subset – T follicular helper (Tfh) cells. We speculate that in patients with OIT-induced EoE, OIT successfully suppresses Tfh cell activity while simultaneously increasing local peTh2 cell activity. This process occurs along a continuum that accounts for the varying manifestations of OIT-related gastrointestinal AEs. At one of the spectrum, Tfh cell-mediated, IgE-predominant symptoms include oropharyngeal pruritis and immediate abdominal pain/vomiting. At the other end of the spectrum peTh2 cell-mediated symptoms include dysphagia, food impaction, and delayed abdominal pain. Some symptoms may be mixed or attributable to either mechanism (e.g., food aversion, nausea, dyspepsia, chest pain, or reflux).

The factors determining whether a patient will develop OIT-induced EoE remain unknown, but emergent disease likely depends on: (1) the degree of epithelial barrier disruption; (2) the magnitude/duration of antigen stimulation, and (3) host factors influencing type 2 immune responses and barrier integrity repair mechanisms. Future research should focus on optimization of antigen administration (e.g., dose, route, and timing) and the genetic and environmental factors that influence epithelial barrier integrity. Ideally, this will facilitate biomarker discovery and enable a precision medicine-based approach to OIT that risk stratifies individuals and tailors interventions accordingly. In addition, more controlled clinical trials are needed to evaluate treatment strategies when OIT-induced complications arise.

Abbreviations/Acronyms:

AE

adverse event

AEC

absolute eosinophil count

DIS

dilated intercellular spaces

EGD

esophagogastroduodenoscopy

EGID

eosinophilic gastrointestinal disease

ELORS

eosinophilic esophagitis-like oral immunotherapy-related syndrome

EoE

eosinophilic esophagitis

eos/hpf

eosinophils per high power field

OITIGERS

oral immunotherapy-induced gastrointestinal symptoms and eosinophilic responses

peTh2 cell

pathogenic effector T helper type 2 cell

Tfh cells

T follicular helper cell