Allergy

REVIEW ARTICLE

Eosinophilic esophagitis is characterized by a non-IgE-

mediated food hypersensitivity§
D. Simon1,*, A. Cianferoni2,3,*, J. M. Spergel2,3, S. Aceves4, M. Holbreich5, C. Venter6,7,
M. E. Rothenberg6, I. Terreehorst8, A. Muraro9, A. J. Lucendo10, A. Schoepfer11, A. Straumann12 &
H.-U. Simon13
1
Department of Dermatology, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland; 2Division of Allergy and
Immunology, Children’s Hospital Philadelphia, University of Pennsylvania; 3Perelman School of Medicine, University of Pennsylvania,
Philadelphia, PA; 4Division of Allergy and Immunology, Department of Pediatrics and Medicine, Center for Infection, Inflammation, and
Immunology, La Jolla, CA; 5Allergy and Asthma Consultants, Indianapolis, IN; 6Division of Allergy and Immunology, Cincinnati Children’s
Hospital Medical Center, Cincinnati, OH, USA; 7School of Health Sciences and Social Work, University of Portsmouth, Portsmouth, UK;
8
Department of ENT and Pediatrics, AMC, Amsterdam, The Netherlands; 9Food Allergy Referral Centre Veneto Region, Padua General
University Hospital, Padua, Italy; 10Department of Gastroenterology, Hospital General de Tomelloso, Tomelloso, Spain; 11Division of
Gastroenterology and Hepatology, Centre Hospitalier Universitaire Vaudois/CHUV, Lausanne; 12Swiss EoE Research Network, Olten;
13
Institute of Pharmacology, University of Bern, Bern, Switzerland

To cite this article: Simon D, Cianferoni A, Spergel JM, Aceves S, Holbreich M, Venter C, Rothenberg ME, Terreehorst I, Muraro A, Lucendo AJ, Schoepfer A,
Straumann A, Simon HU. Eosinophilic esophagitis is characterized by a non-IgE-mediated food hypersensitivity. Allergy 2016; 71: 611–620.

Keywords Abstract
eosinophilic esophagitis; epithelial barrier;
food allergy; immune response;
Eosinophilic esophagitis (EoE) is a chronic disease characterized clinically by
immunoglobulin E. symptoms of esophageal dysfunction and histologically by eosinophil-predomi-
nant inflammation. EoE is frequently associated with concomitant atopic diseases
Correspondence and immunoglobulin E (IgE) sensitization to food allergens in children as well as
Dagmar Simon, MD, Department of to aeroallergens and cross-reactive plant allergen components in adults. Patients
Dermatology, Inselspital, Bern University with EoE respond well to elemental and empirical food elimination diets. Recent
Hospital, CH-3010 Bern, Switzerland. research has, however, indicated that the pathogenesis of EoE is distinct from
Tel.: +41 31 632 2278 IgE-mediated food allergy. In this review, we discuss the individual roles of
Fax: +41 31 632 2233
epithelial barrier defects, dysregulated innate and adaptive immune responses,
E-mail: [email protected]
and of microbiota in the pathogenesis of EoE. Although food has been recog-
§ nized as a trigger factor of EoE, the mechanism by which it initiates or facilitates
Review initiated by the EAACI Eosinophilic
Esophagitis Interest Group eosinophilic inflammation appears to be largely independent of IgE and needs to
*These authors contributed equally to this be further investigated. Understanding the pathogenic role of food in EoE is a
paper. prerequisite for the development of specific diagnostic tools and targeted thera-
peutic procedures.
Accepted for publication 17 January 2016

DOI:10.1111/all.12846

Edited by: Thomas Bieber

Abbreviations
ACD, allergic contact dermatitis; AD, atopic dermatitis; APT, atopy patch test; DHR, drug hypersensitivity reactions; EoE, eosinophilic
esophagitis; FPIES, food protein-induced enterocolitis; GERD, gastroesophageal reflux disease; GI, gastrointestinal; IBD, inflammatory bowel
disease; Ig, immunoglobulin; OAS, oral allergy syndrome; PPI-REE, proton pump inhibitor-responsive esophageal eosinophilia; SFED, six-food
elimination diet; SPT, skin prick test; TCR, T-cell receptor; TNF, tumor necrosis factor.

Allergy 71 (2016) 611–620 © 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 611
EoE: a non-IgE-mediated hypersensitivity Simon et al.

in the pathogenesis? In this review initiated by the EAACI

Current definition of EoE
Eosinophilic Esophagitis Interest Group, we will discuss
As a consequence of intense research in the field of esophageal recently published work on EoE in the context of an
eosinophilia, our understanding of eosinophilic esophagitis immune/antigen-mediated disease.
(EoE) has developed from strict clinic–pathologic criteria lead-
ing toward a conceptual definition which includes pathogenic
EoE-associated IgE sensitization to food and
aspects (1–3). According to current recommendations, EoE
aeroallergens
represents a chronic, immune/antigen-mediated esophageal
disease characterized clinically by symptoms of esophageal EoE is associated with elevated total IgE levels as well as IgE
dysfunction and histologically by eosinophil-predominant sensitization to food and aeroallergens (10). In a pediatric
inflammation (2). EoE has been recognized as having a spec- EoE cohort, sensitizations to food and environmental aller-
trum of clinical signs and symptoms, endoscopic findings as gens have been observed in 75% and 79%, respectively (11).
well as pathologic features. However, the term ‘immune/anti- Skin prick testing in children with EoE revealed increasing
gen-mediated’ does not address the question where EoE should reactivity with inhalant allergen with age, while the reactivity
be positioned in the wide range between autoimmune and aller- to foods decreased (12). Children with EoE were mainly sen-
gic diseases. In addition, there are likely subgroups of patients sitized to milk, eggs, soy, wheat/rye, beef and peanuts (13).
who do not meet this strict definition; for example, some have In adult patients with EoE, specific IgEs to food and inha-
less than 15 eosinophils per high power field (hpf), but other- lant allergen components have been detected in 91% (14).
wise fulfill the criteria of EoE (2). These patients were mainly sensitized to pollens, in particular
The two most common causes of eosinophilia in the esopha- cross-reactive plant allergen components such as profilins
gus, normally devoid of eosinophils in healthy humans, are and pathogenesis-related (PR) 10 proteins (14). Noteworthy
gastroesophageal reflux (GERD) and EoE (2). However, yet is the observation of local immunoglobulin class switching
another form of esophageal eosinophilia has recently emerged and production of IgE in the esophageal mucosa of pediatric
having clinical manifestations and histological features indis- patients with EoE (15). Considering all these findings, EoE
tinguishable from EoE, but distinct from GERD, apart from was initially suspected of being an IgE-mediated allergy to
the fact that it is responsive to high dose of PPI whereas EoE is food and cross-reactive plant allergens.
histologically refractory to PPI. Hence, it is called PPI-respon- On the other hand, clinical trials of targeted food elimina-
sive esophageal eosinophilia (PPI-REE) (4, 5). Patients with tion diets, as well as of IgE blocking, failed to show an IgE-
PPI-REE frequently exhibit environmental and/or food aller- mediated mechanism. Measuring specific IgE levels and/or
gen sensitizations like patients with EoE, whereas the atopy skin prick testing were not sufficient to clearly identify causa-
rate in patients with GERD is similar to that of the general tive food allergens (13, 14, 16, 17). Moreover, elimination
population. Moreover, the inflammatory markers of PPI-REE diets based solely on IgE sensitization to food allergens as
are more similar to those of EoE than of GERD: positive for determined by skin prick tests (SPTs) and/or specific IgE
factors involved in eosinophil chemotaxis (eotaxin-3, CCL26), determinations could not improve EoE in a significant num-
barrier integrity (desmoglein-1, DSG1), tissue remodeling ber of patients (16, 18, 19). The positive predictive values for
(periostin, POSTN), and mast cell-specific activity (car- causative food identified by SPT ranged from 26% to 96%,
boxypeptidase A, CPA3) (4). The molecular signature typical with an average of 47% (16). Based on the assumption that
of PPI-REE and EoE could be reversed by PPI therapy only in IgE plays a key role in pathogenesis, a therapy with an anti-
PPI-REE (4), suggesting the molecular signature is either a IgE antibody for 12 weeks in pediatric and adult patients
sign of disease or marker of eosinophilic inflammation. Mecha- with EoE was initiated in a non-placebo-controlled study
nisms proposed to explain the PPI response include an acid- resulting in a remission rate of only 33% despite an effective
independent, anti-inflammatory action of PPIs on the one reduction of IgE levels observed in the esophageal tissue (20).
hand, or a PPI-induced restoration of esophageal barrier func- In a double-blind placebo-controlled study, anti-IgE treat-
tion on the other (6). In summary, it is possible that PPI-REE ment was not better than placebo in inducing EoE remission
and EoE are the consequence of the same underlying immuno- (21). Taken together, recent clinical and research data lead us
logic mechanism, but additional research is required to confirm to conclude that EoE, while often associated with IgE sensiti-
this concept. zation, is not simply an IgE-mediated food allergy.
Already early reports on EoE mentioned concomitant
allergic diseases and elevated total serum immunoglobulin E
EoE exhibits features of a Th2-predominant
(IgE) levels in about 70% of the patients (7, 8). After receiv-
inflammation
ing elemental formulas, children with esophageal eosinophilia
not responding to pharmacological and/or surgical antireflux The inflammation of EoE is predominantly eosinophilic, but is
therapy, showed marked improvements (9). This observation also characterized by increased numbers of T cells and mast
suggested that EoE could represent an allergic disease in cells infiltrating the esophageal mucosa, as well as high expres-
which food proteins play an important role. However, further sion levels of IL-5 and TNF-a (Fig. 1) (22). Transcriptome
research revealed that EoE seems not to be simply an IgE- analysis of EoE tissue showed a distinct Th2 pattern with sig-
mediated food allergy. What, then, are the underlying causes nificantly elevated mRNA levels of eotaxin-3, IL-5, IL-5 recep-
of EoE and what role might food and/or other antigens play tor a-chain and IL-13 (23, 24). In experimental models, both

612 Allergy 71 (2016) 611–620 © 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Simon et al. EoE: a non-IgE-mediated hypersensitivity

Figure 1 Food as a trigger in the pathogenesis of EoE. (A) In addi- the regional lymph nodes where they stimulate food-specific T
tion to the presence of a genetic predisposition or reflux disease, a cells. Food proteins may induce T-cell responses either as a conse-
food allergy would further disrupt the epithelial barrier and affect quence of antigen presentation by dendritic cells (D) or directly (E)
the microbiota. (B) Food allergens could then penetrate also in the with subsequent eosinophil activation. By releasing toxic granule
skin, bind to pathogen-related receptors and activate epithelial cells proteins and cytokines, eosinophils defend against invading patho-
to produce pro-inflammatory cytokines responsible for the recruit- gens, but cause tissue damage, stimulating fibrosis and perpetuat-
ment and activation of inflammatory cells including eosinophils. (C) ing inflammation. The pathomechanisms of EoE overlapping with
Antigen-presenting cells capturing food antigens would migrate to other diseases are indicated.

eotaxin and IL-5 were essential for eosinophil recruitment, likely contribute to the pathogenesis of EoE, but are not the
accumulation and activation in the esophagus as well as for sole players as pro-inflammatory cytokines are also expressed
epithelial hyperplasia and remodeling (25–28). Moreover, IL- that may regulate additional responses.
13 can induce eotaxin-3 production by esophageal epithelial
cells (29). In addition to the Th2 cytokines, patients with EoE
Lessons learnt from hypersensitivity reactions of the
show elevated blood levels of IL-1a, IL-6 and IL-8, but lower
skin
levels of IL-12, IL-17 and CD40L as compared with healthy
controls, while the gene expression of receptors for IL-1, IL-9 EoE shares many similarities with dermatoses that are due to
and IL-17 is also upregulated in EoE lesions (23, 24). T-cell responses of the skin independent of IgE. Therefore, it
Treatment with corticosteroids resulted in a reduced appears logical to consider antigen-triggered T cell-mediated
expression of eotaxin-3, IL-5 and IL-13 and was followed by mechanisms for the pathogenesis of EoE (Fig. 1).
a decrease of eosinophil numbers in the esophagus of patients
with EoE (29). Although reducing eosinophil inflammation in
T-cell responses in allergic contact dermatitis
the esophagus, blocking IL-5 or IL-13 with therapeutic anti-
bodies has yet to be proven to be clinically useful, although In allergic contact dermatitis (ACD), chemical allergens pene-
trends have been seen in preliminary studies (30, 31). In sum- trate into the skin where they form complexes or bind cova-
mary, Th2 immune responses are a striking feature and most lently to proteins of immune and structural cells in the skin

Allergy 71 (2016) 611–620 © 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 613
EoE: a non-IgE-mediated hypersensitivity Simon et al.

and, thus, may induce innate immune responses as well as diagnosis of food allergy in EoE (16) perhaps owing to the
generate T-cell epitopes (32). Contact allergens, for example, fact that here the skin and not the esophagus is tested. Upon
nickel, are recognized by pattern recognition receptors food allergen, but not nonspecific stimulation, peripheral
(PRRs) resulting in the production of pro-inflammatory blood mononuclear cells from EoE patients with or without
cytokines such as IL-1 and IL-18. This irritant effect of con- allergen-specific IgE produce significant amounts of IL-5
tact allergens is essential for the subsequent activation of the (48). In peanut-allergic children, skin- and gut-homing T cells
adaptive immune system leading to a Tc1/Th1 and Tc17/ expressing Th2 and Th9 genes as well as IL-9 and IL-5 pro-
Th17 effector/memory T-cell response (33). Contact hyper- duction by distinct T helper cell populations have been
sensitivity is dependent on T cell-mediated cytotoxicity via reported (49).
FAS/FASL and perforin pathways (34). In ACD, Th1/Th17 To date, the presence of food allergen-specific T cells in
cells may amplify the cytotoxic cascade as they increase EoE has not been demonstrated. Furthermore, it remains
T cell–keratinocyte adhesiveness and promote ICAM-1- uncertain when and where the sensitization to food allergens
dependent non-antigen-specific keratinocyte killing by T occurs. In adult patients with EoE, airway allergy precedes
lymphocytes (35). However, there is little evidence for an IL- EoE (50). Recent data suggest that an epicutaneous sensitiza-
17-mediated process in EoE (36). tion with ovalbumin may result in an antigen-induced gas-
trointestinal food allergy via the TSLP–basophil axis or in an
IL-17-mediated response depending on the animal model (51,
T cells in drug hypersensitivity
52). Furthermore, filaggrin mutations as risk factors for
While immediate allergic drug hypersensitivity reactions eczema, the atopic march and peanut allergy have been
(DHR) are mediated by specific IgE bound to mast cells and reported, indicating that an impaired epithelial barrier
basophils, delayed (nonimmediate) allergic DHR are T cell- function may predispose to allergen sensitization and atopy
mediated. Analogous to haptens, drugs are presented either (53, 54).
covalently bound to peptides in the binding grove of MHC
molecules on antigen-presenting cells or complexed to amino
Epithelial barrier and innate immune responses in EoE
acids in MHC molecules and TCR (37). Recently, a concept
for the pharmacological interaction of drugs with immune There is increasing evidence that EoE is associated with a
receptor (p-i concept) has been proposed, suggesting a non- dysfunction at the epithelial barrier followed by an eosino-
covalent binding enabling a direct interaction with immuno- philic inflammation similar to AD which is concomitant in
logical receptors such as MHC and TCR (38, 39). Thus, the over half of patients with EoE (Fig. 1). In esophageal epithe-
antigen might bind either to the MHC complex, thereby lial cells, the expression of epidermal differentiation complex
modifying the structure that is recognized by the TCR lead- (EDC) genes, for example, filaggrin, SPRR3 and keratins, is
ing to a specific T-cell activation, or directly to a specific downregulated in response to IL-13 and in active EoE, where
TCR requiring additional MHC interaction for full T-cell it could be only partially normalized upon therapy (55, 56).
activation (38). In cutaneous reactions, drug-specific cyto- Desmoglein (DSG)-1, an intercellular adhesion molecule
toxic T cells have been demonstrated that can contribute to responsible for epithelial integrity and barrier function was
tissue damage via perforin/granzyme B or FAS/FASL mech- one of the most strongly downregulated genes in EoE (29). A
anisms (40, 41). In DRESS, an oligoclonal expansion of acti- downregulation of DSG-1 gene, for example, by IL-13, was
vated CD8+ T cells directed against viral antigens derived shown to result in the separation of epithelial cells (spongio-
from Herpes viruses, whose replication is enhanced by the sis) followed by impaired barrier function as well as by peri-
culprit drug, has been observed in the skin and visceral ostin induction further potentiating inflammation (57).
organs (42). Ultrastructural analysis revealed a significantly decreased
number of desmosomes per cell in EoE biopsies as compared
to healthy controls, which was reversible after treatment (58).
Food-specific T-cell responses in the skin
Furthermore, the expression of filaggrin and the tight junc-
Over 80% of patients with atopic dermatitis (AD) have tion proteins zonula occludens (ZO)-3 and claudin-1 is
increased IgEs to foods and inhalant allergens in the periph- decreased in EoE, correlating with spongiosis (59). Consistent
eral blood (43). However, the positive predictive value of IgE with this finding, mutations in filaggrin are overrepresen-
specific to food allergens is low (44). Interestingly, in 45% of ted in patients with EoE (55) and homozygous mutations
patients reacting upon food allergen challenge, eczematous of DSG1 cause a severe atopy syndrome which includes
reactions with or without prior immediate reactions have EoE (60).
been observed, suggesting the occurrence of late, most likely In stratified epithelia, the activity of proteases is tightly
T cell-mediated reactions against foods (44). Indeed, in regulated by protease inhibitors. The loss of inhibition results
patients with food-triggered AD exacerbations, relevant food in cleavage of desmosomal proteins and loss of barrier integ-
allergen-specific T cells have been detected in the peripheral rity, facilitating the penetration of allergens and microbes as
blood as well as the skin (45, 46). Moreover, positive atopy well as the subsequent generation of danger signals and pro-
patch test (APT) reactions to inhalant and food allergens can tease activated receptor (PAR)-2 activation (61). In active
be detected in the absence of corresponding IgE responses EoE, a significantly decreased expression of the protease inhi-
(47). Although widely used, the APT has limited value in the bitor LEKTI has been observed (36).

614 Allergy 71 (2016) 611–620 © 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Simon et al. EoE: a non-IgE-mediated hypersensitivity

TSLP that is produced by epithelial cells in response to to barrier function in UC (72). Due to an increased intestinal
PAR-2, Toll-like receptor (TLR) stimulation or mechanical epithelial permeability, food antigens and microbes may acti-
injury, strongly induces Th2 immune responses by stimulat- vate pattern recognition receptors on epithelial cells resulting
ing dendritic cells, T cells, eosinophils, mast cells and baso- in a release of pro-inflammatory cytokines such as TNF-a,
phils (62). Upon stimulation with TSLP, eosinophils that IL-1, IL-18 and IL-33 (Fig. 1) (71). In contrast to EoE, pre-
bear the TSLP receptor on their surface generate extracellular dominantly Th1 cells and the IL-23/Th17 axis are activated
DNA traps associated with granule proteins that are able to in IBD (71). It has been hypothesized that due to a dysregu-
kill bacteria (63). Interestingly, the expression of TSLP is lated innate intestinal immunity and barrier function, affect-
increased in EoE and correlates with the number of eosino- ing both the diversity and composition of the microbiota, the
phils generating eosinophil extracellular traps (36). Genetic immune response is initiated to eliminate invading antigens
variants of TSLP and its receptor have been associated with (e.g. microbes, food) and to restore epithelial barrier integ-
an increased susceptibility to EoE overall, and in males, rity, but may later turn into a chronic inflammation leading
respectively (64, 65). Furthermore, the gene of esophageal to the clinical manifestations of IBD (71, 73).
selective calpain (CAPN) 14, a member of the calpain pro- Thus, the principal pathomechanisms of IBD seem congru-
tease family involved in the cleavage of inflammatory media- ent with those of EoE, although it is currently not clear which
tors such as IL-33, was upregulated in active EoE, while the tissue-specific characteristics, including immune responses,
calpain inhibitor CAST was downregulated (66). In line with environmental factors such as microbiota and food, as well as
these findings, genetic variants in the CAPN 14 gene locus genetic predispositions favor a chronic Th1/Th17 inflamma-
are linked with EoE susceptibility (67) and increased expres- tion as in IBD or a Th2-predominant inflammation as in EoE
sion of innate cytokines including IL-33 by epithelial cells with corresponding clinical phenotypes. While both diseases
has been detected in EoE (36). have common mechanisms, the upstream events are likely to
Immense efforts have been undertaken to identify the role be different as EoE is associated with unique genetic suscepti-
of the microbiota in the immune system, in particular in bility (TSLP and CAPN14) and atopy; whereas IBD is more
association with immune-mediated diseases. Microbiota related to innate immunity to microbial flora.
research aims at elucidating their role in initiating and per-
petuating inflammation and, conversely, the effect of diseases
EoE is distinct from IgE-mediated food allergies
and treatment procedures on the microbiota. Compared to
healthy controls, the bacterial load of the esophagus is If one were to consider EoE as a kind of food allergy, how
increased in patients with EoE regardless of treatment and would its symptoms agree with the current concept of gas-
disease activity, with a relative abundance of gram-negative trointestinal (GI) allergies? A food allergy is defined as an
bacteria in active EoE (68, 69). Recently, IgE sensitization to abnormal immunologic response to a food substance occur-
Candida albicans has been reported in pediatric and adult ring in a susceptible host and causing some type of GI
patients with EoE (14, 70). Whether an esophageal coloniza- inflammation. The vast majority of food allergies affecting
tion with Candida albicans and later sensitization is owing to the GI tract are characterized by a Th2 inflammation with
EoE inflammation or corticosteroid therapy remains to be predominant Th2 cytokine expression (that is IL-4, IL-13,
investigated. Furthermore, any potential role of IgE specific and IL-5). Th2 inflammation can cause B cells to produce
for Candida albicans in the pathogenesis of EoE is uncertain. IgE antibodies specific to certain foods or can lead to a
Taken together, recent research suggests that impaired chronic cellular inflammation frequently characterized by the
epithelial barrier function plays a major role in initiating and presence of Th2 cell and eosinophils (74).
perpetuating EoE inflammation as it facilitates the penetra- According to the immunological mechanism elicited, food
tion of allergens and microbes and generates danger signals allergies can be classified into (1) IgE-mediated, which are
leading to an activation of epithelial cells as well as innate immediate, short-lived reactions mediated by antibodies
and adaptive immune cells with subsequent chemokine and belonging to the IgE class; (2) cell-mediated, which usually
cytokine production resulting in Th2 immune responses. have a delayed/chronic course, typically involving the GI tract
There is evidence of a dysbiosis of microbiota in EoE; how- and the cell component of the immune system responsible for
ever, the consequences in terms of microbial-triggered eosino- inflammation; or (3) mixed, IgE- and cell-mediated (75). IgE-
philic inflammation and the particular role of diet on the mediated reactions to foods are acute and highly reproducible.
microbiome in the esophagus remain to be investigated. They are initiated by the cross-linking of two or more allergen-
specific IgE antibodies bound to their high-affinity receptor
(FceRI) expressed on mast cells and basophils as a result of a
Similarities and differences between EoE and IBD
specific food allergen engagement. Such cross-linking determi-
With inflammatory bowel diseases (IBDs), such as Crohn’s nes the release of preformed mediators, in particular, his-
disease (CD) and ulcerative colitis (UC), the pathogenesis is tamine, that cause vasodilatation, angioedema, smooth muscle
determined by genetic factors, environmental and microbial constriction, and increased mucus production (76).
factors together with an epithelial barrier dysfunction and Examples of typical IgE-mediated allergic reactions affect-
subsequent innate and adaptive immune responses (71). The ing the GI tract are the oral allergy syndrome (OAS) and the
susceptibility to IBD is determined by genetic variants related more severe GI food allergy, also known as ‘gastrointestinal
to innate immunity, autophagy and phagocytosis in CD and anaphylaxis’. When comparing IgE-mediated OAS and GI

Allergy 71 (2016) 611–620 © 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 615
EoE: a non-IgE-mediated hypersensitivity Simon et al.

food allergies with EoE, the following differences become evi- the antifood IgG4 levels did not correlate with the age and
dent: EoE symptoms might be instant, but they are not tran- duration of disease symptoms (21). Further studies will be
sient, EoE inflammation is chronic, anaphylaxis is not a necessary to really understand the pathogenic role of IgG4
feature of EoE, and pollen-associated food allergens are not in EoE.
a typical trigger of EoE. It should be noted, however, that
patients with EoE can concurrently suffer from OAS and/or
EoE is characterized by a non-IgE-mediated food
a GI food allergy.
hypersensitivity
Food protein-induced enterocolitis (FPIES), an increas-
ingly recognized form of non-IgE-mediated food hypersensi- Since the first description of a series of clinical cases of EoE,
tivity, is characterized by a delayed onset of vomiting with or food allergies have appeared to play a major role in causing
without diarrhea, typically occurring in infants and toddlers a severe esophageal eosinophilia that resolved on elemental
from 2 to 6 h postingestion of the trigger food (77, 78). diet, but not on aggressive GERD treatment, including Nis-
FPIES is usually a transient disease which starts at sen fundoplication (6). In view of this, food allergens have
4–9 months of life or when solid foods are first introduced, been identified as triggers of EoE in most children and adults
and resolves by age 2–5 years (77). The foods most com- (6, 16, 82, 83).
monly involved in FPIES are milk, soy, rice, oats and eggs. Thus, food as a trigger of EoE fulfills Koch’s postulates as
IgEs specific to the trigger foods are usually not detectable the addition or subtraction of foods can cause disease or
(77, 79). Although FPIES and EoE seem to share some clini- eliminate EoE in nearly all patients. The most effective treat-
cal (symptoms, age of onset) and pathogenic (causative food ment in patients with EoE is an elemental diet that induces
triggers, increased TNF-a, epithelial barrier defects) features histological and clinical resolution in over 95% of pediatric
(80), other characteristics such as disease course, endoscopic and adult patients (83–86). Noteworthy is that IBD may also
and histologic findings discriminate FPIES from EoE. resolve upon elemental diet (87) with a mechanism that
involves both bowel rest and a change in microbiome. So
far, the explanation for remission of EoE on an elemental
Experience with omalizumab: Its lack of clinical
diet has always been linked to the avoidance of food aller-
efficacy in EoE
gens, rather than bowel rest/change in microbiome, but this
Omalizumab is an anti-IgE humanized monoclonal antibody possibility needs to be investigated further. This presumption
that binds to the fragment crystallizable (Fc) region of the was supported by the fact that elimination diets based on
IgE molecule and thus prevents its binding to the high-affi- removal of the six most common food allergens (SFED –
nity IgE receptor (Fc epsilon RI, FceRI). In the only pub- six-food elimination diet) (82) or of the foods to which
lished prospective, randomized, double-blind placebo- patients were sensitized (targeted elimination diets) have been
controlled study in 30 adult patients with EoE (16 treated shown to induce and maintain EoE remission in 72% and
with omalizumab and 14 with placebo) omalizumab was 45% of patients with EoE, respectively (13, 16). According
given every 2–4 weeks for 16 weeks, based on weight and to biopsy confirmation, the most common food proteins
serum level of IgE. Before starting the treatment and at the causing EoE are milk, followed by wheat, eggs, beef, soy
end of the trial (16 weeks of treatment) symptoms evalua- and legumes, and chicken (16, 83, 88, 89). Interestingly, pea-
tion, EGD and histological assessment of the eosinophil nuts, tree nuts, fish and shellfish are rare as causes for EoE
density (peak eos/hpf) in esophageal biopsies were per- despite being common causes of IgE-mediated reactions in
formed. Patients treated with omalizumab had neither a sig- adults.
nificant improvement in symptoms nor a decrease of the The evidence that EoE is generally non-IgE-mediated is
eosinophil infiltration of the esophageal mucosa compared based on both clinical and research findings:
with placebo (21). This study confirmed anecdotal data from 1 Despite the fact that the majority of patients with EoE
clinical cases reported in which omalizumab had been con- have specific IgEs to food allergens and/or aeroallergens,
sidered to improve IgE-mediated symptoms of food allergy, the detection of specific IgEs for food allergens, either by
but not of EoE (81). Overall, these data support the notion SPT or by specific sera IgE (sIgE), has not proven suc-
that EoE is not IgE-mediated. Clayton et al. (21) speculated cessful for the identification of causative foods in EoE
that IgG4 antibodies specific for a food allergen are block- (84, 85). Indeed, removal of SPT- or sIgE-positive foods
ing IgE responses. Indeed, in allergic diseases, an IgG4 is not superior to SFED (2, 16, 17, 83, 90). Moreover, it
response follows an IgE-mediated response and does block has been reported that the introduction of skin test-nega-
IgE-mediated mast cell activation (21). In EoE, extracellular tive foods into the diet sometimes induces clinical disease
granular deposits of IgG4 and abundant IgG4-containing (6, 16).
plasma cells in the tissue, as well as increased serum levels 2 Clinical trials and case series have shown that therapy
of IgG4 reactive with specific foods have been observed, with omalizumab is not effective in inducing remission of
suggesting that in adults, EoE might be an IgG4- and IgE- EoE (21, 81).
associated disease and perhaps the balance between the two 3 Oral immunotherapy, which has been used successfully in
antibodies could be a key determinant (21). However, B IgE-mediated food allergy, is associated with an increased
cell-deficient mice also develop typical EoE, suggesting that risk of developing EoE (e.g. in 2 to 10% of treated
antibodies may simply be nonpathogenic (82). Moreover, patients) (91–93).

616 Allergy 71 (2016) 611–620 © 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Simon et al. EoE: a non-IgE-mediated hypersensitivity

4 Children who outgrow IgE-mediated food allergy and

Conclusions
therefore are able to reintroduce these foods in their diet
can later develop EoE to the same food (94). There is strong evidence that foods, most likely food pro-
5 In experimental models in which food allergens are able teins, are triggers of EoE, as elimination of culprit food cate-
to induce an EoE-like disease, mice with depleted IgE gories as well as protein-free elemental diets results in an
and devoid of mast cells still could develop esophageal improvement of histological and endoscopic signs as well as
inflammation and consequent food impaction similar to of symptoms. Furthermore, the observation that the eosino-
the wild-type mice (95, 96). philic inflammation and the Th2 inflammation pattern reap-
pear rapidly after reintroduction of the culprit foods is a
strong evidence that EoE is likely a food-driven disorder with
Confirmation of EoE diagnosis and the practical search
features of food allergy. However, the spectrum of clinical
for offending foods
presentations of EoE, the results of IgE-based diagnostic pro-
EoE is a clinicopathological diagnosis. However, EoE and cedures, and the lack of efficacy of anti-IgE treatment sug-
GERD have a substantial overlap of clinical and of histo- gest that EoE cannot be regarded as an IgE-mediated food
logical features. For instance, the presence of heartburn and allergy.
marked esophageal eosinophilia might be fairly common in The mechanism by which food elicits EoE is not yet under-
both entities (2). To solve this diagnostic conundrum, stood. It seems likely that a cellular mechanism similar to
updated consensus recommendations for diagnosis and man- contact allergy of the skin or drug hypersensitivity plays a
agement of EoE advocate performing a PPI trial in patients role. IgG4 formed against foods has been suspected of play-
having symptoms suggestive of EoE and esophageal eosino- ing a role in EoE, perhaps as a blocking antibody. Analogous
philia (2). Accordingly, a diagnosis of GERD was recom- to AD and IBD, an impairment of the epithelial barrier,
mended for those patients responding to PPI therapy, alterations of the microbiota and subsequent chronic inflam-
whereas patients whose symptoms and inflammation persist mation might be the underlying pathogenic factors for EoE.
were regarded as having EoE (2). Unfortunately, this diag- Given this scenario, food might interfere either as an irritant,
nostic PPI trial did not fulfill the expectation of differentiat- modulator of the microbiota or as an antigen/allergen to ini-
ing EoE from GERD, but unexpectedly uncovered a third tiate and perpetuate inflammation (Fig. 1). The identification
category of patients, called PPI-REE, presenting with symp- of offending foods by empirical elimination diets and con-
toms of EoE, but responding to PPI (5). With the exception trolled reintroduction of foods is inconvenient for patients,
of the responsiveness to PPI, PPI-REE, and EoE have time-consuming, and in the clinical routine hardly applicable.
common clinical, endoscopic, histological and molecular Nevertheless, this procedure is currently the only reliable
features. method to identify food triggers in patients with EoE. Eluci-
EoE is a chronic and progressive disease. If left dating the exact mechanism of how foods affect EoE would
untreated complications, such as food impaction, esopha- allow the development of novel diagnostic tests. For instance,
geal stricture, narrow-caliber esophagus, and esophageal the determination of food-specific markers including T-cell
perforation, are common (97, 98). Therefore, once the diag- responses to specific foods could possibly overcome the limi-
nosis is confirmed, it is important to treat the eosinophilic tations of SPT, APT, and empirical diets. As in IBD and ato-
inflammation not only to control the presenting symptoms, pic diseases, EoE should be considered as a complex disease
but also to preserve the morphological and functional with a disordered interplay between the epithelial barrier,
integrity of the esophagus (2, 10, 87, 98). Beside medica- innate and adaptive immune responses together with the
tions, diets avoiding culprit foods are an important thera- composition of the microbiota.
peutic option (99). Of note, before an elimination diet can
be established, it is necessary to identify the triggering
Conflict of interest
foods, ideally with the help of a dietitian specialized in
dealing with this disease. Currently culprit foods are identi- J. M. Spergel has consultant contracts with DBV Technology
fied by demonstrating histological and clinical remission of and Danone and received a grant from Aimmune Therapeu-
EoE after the establishment of an elimination diet. In prac- tics. M. E. Rothenberg is a consultant for Receptos and
tice, after avoidance and after reintroduction of any food NKT Therapeutics, and has an equity interest in Immune
category, the effect must be controlled endoscopically and Pharmaceuticals and NKT Therapeutics. He has received
histologically (82, 83). Serial endoscopies are therefore speaking honorarium from Merck. He is an inventor of eosi-
required to figure out an individual elimination diet. This nophilic esophagitis-related patents owned by Cincinnati
approach is time-consuming, inconvenient for patients, Children’s Hospital Medical Center, some of which have
expensive, and affects the quality of life (100). Therefore, been licensed to Miraca Life Sciences. He has a royalty inter-
there is a need to develop noninvasive methods for the est in reslizumab, a drug being developed by Texas Pharma-
identification of the offending foods. The determination of ceuticals. A. Schoepfer has consultant contracts with Falk,
food-specific IgG4 in the serum is a method currently Receptos, and Regeneron. A. Straumann has consultant con-
under evaluation. Further phenotyping patients based on tracts with Actelion, Falk, Genentech-Roche, Nutricia,
their esophageal gene expression, using a 94 gene transcript Receptos, and Regeneron. The remaining authors have no
profile, is promising to be helpful (101). potential conflict of interest.

Allergy 71 (2016) 611–620 © 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 617
EoE: a non-IgE-mediated hypersensitivity Simon et al.

References
1. Furuta GT, Liacouras CA, Collins MH, sensitization in eosinophilic esophagitis. A et al. A striking local esophageal cyto-
Gupta SK, Justinich C, Putnam PE et al. Allergy 2007;62:1257–1260. kine expression profile in eosinophilic
Eosinophilic esophagitis in children and 13. Spergel JM, Beausoleil JL, Mascarenhas M, esophagitis. J Allergy Clin Immunol
adults: a systematic review and consensus Liacouras CA. The use of skin prick tests 2011;127:208–217.
recommendations for diagnosis and and patch tests to identify causative foods 25. Mishra A, Hogan SP, Brandt EB, Rothen-
treatment. Gastroenterology 2007;133:1342– in eosinophilic esophagitis. J Allergy Clin berg ME. An etiological role for aeroaller-
1363. Immunol 2002;109:363–368. gens and eosinophils in experimental
2. Liacouras CA, Furuta GT, Hirano I, 14. Simon D, Straumann A, Dahinden C, esophagitis. J Clin Invest 2001;107:83–90.
Atkins D, Attwood SE, Bonis PA et al. Simon HU. Frequent sensitization to Can- 26. Mishra A, Rothenberg ME. Intratracheal
Eosinophilic esophagitis: updated consensus dida albicans and profilins in adult eosino- IL-13 induces eosinophilic esophagitis by
recommendations for children and adults. philic esophagitis. Allergy 2013;68:945–948. an IL-5, eotaxin-1, and STAT6-dependent
J Allergy Clin Immunol 2011;128:3–20. 15. Vicario M, Blanchard C, Stringer KF, Col- mechanism. Gastroenterology
3. Furuta GT, Katzka DA. Eosinophilic lins MH, Mingler MK, Ahrens A et al. 2003;125:1419–1427.
esophagitis. N Engl J Med 2015;373: Local B cells and IgE production in the 27. Kohan M, Puxeddu I, Reich R, Levi-Schaf-
1640–1648. oesophageal mucosa in eosinophilic fer F, Berkman N. Eotaxin-2/CCL24 and
4. Wen T, Dellon ES, Moawad FJ, Furuta oesophagitis. Gut 2010;59:12–20. eotaxin-3/CCL26 exert differential profibro-
GT, Aceves SS, Rothenberg ME. Tran- 16. Spergel JM, Brown-Whitehorn TF, Cianfer- genic effects on human lung fibroblasts.
scriptome analysis of proton pump inhibi- oni A, Shuker M, Wang ML, Verma R Ann Allergy Asthma Immunol 2010;104:
tor-responsive esophageal eosinophilia et al. Identification of causative foods in 66–72.
reveals proton pump inhibitor-reversible children with eosinophilic esophagitis trea- 28. Zuo L, Fulkerson PC, Finkelman FD,
allergic inflammation. J Allergy Clin Immu- ted with an elimination diet. J Allergy Clin Mingler M, Fischetti CA, Blanchard C
nol 2015;135:187–197. Immunol 2012;130:461–467. et al. IL-13 induces esophageal remodeling
5. Molina-Infante J, Bredenoord AJ, Cheng 17. Henderson CJ, Abonia JP, King EC, Put- and gene expression by an eosinophil-inde-
E, Dellon ES, Furuta GT, Gupta SK et al. nam PE, Collins MH, Franciosi JP et al. pendent, IL-13R alpha 2-inhibited pathway.
Proton pump inhibitor-responsive oesopha- Comparative dietary therapy effectiveness J Immunol 2010;185:660–669.
geal eosinophilia: an entity challenging in remission of pediatric eosinophilic 29. Blanchard C, Mingler MK, Vicario M,
current diagnostic criteria for eosinophilic esophagitis. J Allergy Clin Immunol Abonia JP, Wu YY, Lu TX et al. IL-13
oesophagitis. Gut 2015;65:524–531. 2012;129:1570–1578. involvement in eosinophilic esophagitis:
6. Eluri S, Dellon ES. Proton pump inhibitor- 18. Simon D, Straumann A, Wenk A, Spichtin transcriptome analysis and reversibility with
responsive oesophageal eosinophilia and H, Simon HU, Braathen LR. Eosinophilic glucocorticoids. J Allergy Clin Immunol
eosinophilic oesophagitis: more similarities esophagitis in adults–no clinical relevance 2007;120:1292–1300.
than differences. Curr Opin Gastroenterol of wheat and rye sensitizations. Allergy 30. Straumann A, Conus S, Grzonka P, Kita
2015;31:309–315. 2006;61:1480–1483. H, Kephart G, Bussmann C et al. Anti-
7. Attwood SE, Smyrk TC, Demeester TR, 19. Spergel JM, Andrews T, Brown-Whitehorn interleukin-5 antibody treatment (me-
Jones JB. Esophageal eosinophilia with TF, Beausoleil JL, Liacouras CA. Treat- polizumab) in active eosinophilic
dysphagia, a distinct clinicopathologic syn- ment of eosinophilic esophagitis with speci- oesophagitis: a randomised, placebo-con-
drome. Dig Dis Sci 1993;38:109–116. fic food elimination diet directed by a trolled, double-blind trial. Gut 2010;59:
8. Straumann A, Spichtin HP, Bernoulli R, combination of skin prick and patch tests. 21–30.
Loosli J, V€ ogtlin J. Idiopathic eosinophilic Ann Allergy Asthma Immunol 2005;95: 31. Rothenberg ME, Wen T, Greenberg A,
esophagitis: a frequently overlooked disease 336–343. Alpan O, Enav B, Hirano I et al. Intra-
with typical clinical aspects and discrete 20. Loizou D, Enav B, Komlodi-Pasztor E, venous anti-IL-13 mAb QAX576 for the
endoscopic findings. Schweiz Med Wochen- Hider P, Kim-Chang J, Noonan L et al. treatment of eosinophilic esophagitis. J
schr 1994;124:1419–1429. A pilot study of omalizumab in eosinophi- Allergy Clin Immunol 2015;135:500–507.
9. Kelly KJ, Lazenby AJ, Rowe PC, Yardley lic esophagitis. PLoS One 2015;10: 32. Martin SF. Allergic contact dermatitis:
JH, Perman JA, Sampson HA. Eosinophilic e0113483. xenoinflammation of the skin. Curr Opin
esophagitis attributed to gastroesophageal 21. Clayton F, Fang JC, Gleich GJ, Lucendo Immunol 2012;24:720–729.
reflux: improvement with an amino acid- AJ, Olalla JM, Vinson LA et al. Eosinophi- 33. Martin SF. New concepts in cutaneous
based formula. Gastroenterology lic esophagitis in adults is associated with allergy. Contact Dermatitis 2015;72:2–10.
1995;109:1503–1512. IgG4 and not mediated by IgE. Gastroen- 34. Kehren J, Desvignes C, Krasteva M,
10. Straumann A, Aceves SS, Blanchard C, terology 2014;147:602–609. Ducluzeau MT, Assossou O, Horand F
Collins MH, Furuta GT, Hirano I et al. 22. Straumann A, Bauer M, Fischer B, Blaser et al. Cytotoxicity is mandatory for CD8(+)
Pediatric and adult eosinophilic esophagitis: K, Simon HU. Idiopathic eosinophilic T cell-mediated contact hypersensitivity.
similarities and differences. Allergy esophagitis is associated with a T(H)2-type J Exp Med 1999;189:779–786.
2012;67:477–490. allergic inflammatory response. J Allergy 35. Pennino D, Eyerich K, Scarponi C, Car-
11. Assa’ad AH, Putnam PE, Collins MH, Clin Immunol 2001;108:954–961. bone T, Eyerich S, Nasorri F et al. IL-17
Akers RM, Jameson SC, Kirby CL et al. 23. Blanchard C, Wang N, Stringer KF, Mis- amplifies human contact hypersensitivity by
Pediatric patients with eosinophilic hra A, Fulkerson PC, Abonia JP et al. licensing hapten nonspecific Th1 cells to kill
esophagitis: an 8-year follow-up. J Allergy Eotaxin-3 and a uniquely conserved gene- autologous keratinocytes. J Immunol
Clin Immunol 2007;119:731–738. expression profile in eosinophilic esophagi- 2010;184:4880–4888.
12. Sugnanam KK, Collins JT, Smith PK, tis. J Clin Invest 2006;116:536–547. 36. Simon D, Radonjic-H€ osli S, Straumann A,
Connor F, Lewindon P, Cleghorn G et al. 24. Blanchard C, Stucke EM, Rodriguez-Jime- Yousefi S, Simon HU. Active eosinophilic
Dichotomy of food and inhalant allergen nez B, Burwinkel K, Collins MH, Ahrens esophagitis is characterized by epithelial

618 Allergy 71 (2016) 611–620 © 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Simon et al. EoE: a non-IgE-mediated hypersensitivity

barrier defects and eosinophil extracellular duction in adult patients with eosinophilic and spongiosis in esophageal epithelia of
trap formation. Allergy 2015;70:443–452. esophagitis. Dig Dis Sci 2006;51:1934–1941. patients with eosinophilic esophagitis. Clin
37. Martin SF, Esser PR, Schmucker S, Dietz 49. Brough HA, Cousins DJ, Munteanu A, Gastroenterol Hepatol 2014;12:1824–1829.
L, Naisbitt DJ, Park BK et al. T-cell recog- Wong YF, Sudra A, Makinson K et al. IL- 60. Samuelov L, Sarig O, Harmon RM, Rapa-
nition of chemicals, protein allergens and 9 is a key component of memory Th cell port D, Ishida-Yamamoto A, Isakov O et al.
drugs: towards the development of in vitro peanut-specific responses from children Desmoglein 1 deficiency results in severe
assays. Cell Mol Life Sci 2010;67:4171– with peanut allergy. J Allergy Clin Immunol dermatitis, multiple allergies and
4184. 2014;134:1329–1338. metabolic wasting. Nat Genet 2013;45:
38. Pichler WJ, Naisbitt DJ, Park BK. Immune 50. Simon D, Marti H, Heer P, Simon HU, 1244–1248.
pathomechanism of drug hypersensitivity Braathen LR, Straumann A. Eosinophilic 61. Hovnanian A. Netherton syndrome: skin
reactions. J Allergy Clin Immunol esophagitis is frequently associated with inflammation and allergy by loss of pro-
2011;127:74–81. IgE-mediated allergic airway diseases. J tease inhibition. Cell Tissue Res
39. Demoly P, Adkinson NF, Brockow K, Cas- Allergy Clin Immunol 2005;115:1090–1092. 2013;351:289–300.
tells M, Chiriac AM, Greenberger PA et al. 51. Noti M, Kim BS, Siracusa MC, Rak GD, 62. Roan F, Bell BD, Stoklasek TA, Kitajima
International Consensus on drug allergy. Kubo M, Moghaddam AE et al. Exposure M, Han H, Ziegler SF. The multiple facets
Allergy 2014;69:420–437. to food allergens through inflamed skin of thymic stromal lymphopoietin (TSLP)
40. Yawalkar N, Egli F, Hari Y, Nievergelt H, promotes intestinal food allergy through during allergic inflammation and beyond.
Braathen LR, Pichler WJ. Infiltration of the thymic stromal lymphopoietin-basophil J Leukoc Biol 2012;91:877–886.
cytotoxic T cells in drug-induced cutaneous axis. J Allergy Clin Immunol 63. Morshed M, Yousefi S, St€ ockle C, Simon
eruptions. Clin Exp Allergy 2000;30:847– 2014;133:1390–1399. HU, Simon D. Thymic stromal lymphopoi-
855. 52. He R, Kim HY, Yoon J, Oyoshi MK, etin stimulates the formation of eosinophil
41. Schmid S, Kuechler PC, Britschgi M, Stei- MacGinnitie A, Goya S et al. Exaggerated extracellular traps. Allergy 2012;67:1127–
ner UC, Yawalkar N, Limat A et al. Acute IL-17 response to epicutaneous sensitization 1137.
generalized exanthematous pustulosis: role mediates airway inflammation in the 64. Rothenberg ME, Spergel JM, Sherrill JD,
of cytotoxic T cells in pustule formation. absence of IL-4 and IL-13. J Allergy Clin Annaiah K, Martin LJ, Cianferoni A et al.
Am J Pathol 2002;161:2079–2086. Immunol 2009;124:761–770. Common variants at 5q22 associate with
42. Picard D, Janela B, Descamps V, D’Incan 53. Marenholz I, Nickel R, R€ uschendorf F, pediatric eosinophilic esophagitis. Nat
M, Courville P, Jacquot S et al. Drug reac- Schulz F, Esparza-Gordillo J, Kerscher T Genet 2010;42:289–291.
tion with eosinophilia and systemic symp- et al. Filaggrin loss-of-function mutations 65. Sherrill JD, Gao PS, Stucke EM, Blan-
toms (DRESS): a multiorgan antiviral T predispose to phenotypes involved in the chard C, Collins MH, Putnam PE et al.
cell response. Sci Transl Med 2010;2:46–62. atopic march. J Allergy Clin Immunol Variants of thymic stromal lymphopoietin
43. Akdis CA, Akdis M, Simon D, Dibbert B, 2006;118:866–871. and its receptor associate with eosinophilic
Weber M, Gratzl S et al. T cells and T cell- 54. Brown SJ, Asai Y, Cordell HJ, Campbell esophagitis. J Allergy Clin Immunol
derived cytokines as pathogenic factors in LE, Zhao Y, Liao H et al. Loss-of-function 2010;126:160–165.
the nonallergic form of atopic dermatitis. variants in the filaggrin gene are a signifi- 66. Kottyan LC, Davis BP, Sherrill JD, Liu K,
J Invest Dermatol 1999;113:628–634. cant risk factor for peanut allergy. J Rochman M, Kaufman K et al. Genome-
44. Breuer K, Heratizadeh A, Wulf A, Bau- Allergy Clin Immunol 2011;127:661–667. wide association analysis of eosinophilic
mann U, Constien A, Tetau D et al. Late 55. Blanchard C, Stucke EM, Burwinkel K, esophagitis provides insight into the tissue
eczematous reactions to food in children Caldwell JM, Collins MH, Ahrens A et al. specificity of this allergic disease. Nat Genet
with atopic dermatitis. Clin Exp Allergy Coordinate interaction between IL-13 and 2014;46:895–900.
2004;34:817–824. epithelial differentiation cluster genes in 67. Sleiman PM, Wang ML, Cianferoni A,
45. Reekers R, Beyer K, Niggemann B, Wahn eosinophilic esophagitis. J Immunol Aceves S, Gonsalves N, Nadeau K et al.
U, Freihorst J, Kapp A et al. The role of 2010;184:4033–4041. GWAS identifies four novel
circulating food antigen-specific lympho- 56. Kiran KC, Rothenberg ME, Sherrill JD. In eosinophilic esophagitis loci. Nat Commun
cytes in food allergic children with atopic vitro model for studying esophageal epithe- 2014;5:5593.
dermatitis. Br J Dermatol 1996;135:935– lial differentiation and allergic inflamma- 68. Harris JK, Fang R, Wagner BD, Choe
941. tory responses identifies keratin HN, Kelly CJ, Schroeder S et al. Esopha-
46. Reekers R, Busche M, Wittmann M, Kapp involvement in eosinophilic esophagitis. geal microbiome in eosinophilic esophagitis.
A, Werfel T. Birch pollen-related foods PLoS One 2015;10:e0127755. PLoS One 2015;10:e0128346.
trigger atopic dermatitis in patients with 57. Sherrill JD, Kc K, Wu D, Djukic Z, Caldwell 69. Benitez AJ, Hoffmann C, Muir AB, Dods
specific cutaneous T-cell responses to birch JM, Stucke EM et al. Desmoglein-1 regulates KK, Spergel JM, Bushman FD et al.
pollen antigens. J Allergy Clin Immunol esophageal epithelial barrier function and Inflammation-associated microbiota in
1999;104:466–472. immune responses in eosinophilic esophagi- pediatric eosinophilic esophagitis. Micro-
47. Darsow U, Laifaoui J, Kerschenlohr K, tis. Mucosal Immunol 2014;7:718–729. biome 2015;3:23.
Wollenberg A, Przybilla B, W€ uthrich B 58. Capocelli KE, Fernando SD, Menard- 70. Erwin EA, James HR, Gutekunst HM,
et al. The prevalence of positive reactions Katcher C, Furuta GT, Masterson JC, Russo JM, Kelleher KJ, Platts-Mills TA.
in the atopy patch test with aeroallergens Wartchow EP. Ultrastructural features of Serum IgE measurement and detection of
and food allergens in subjects with atopic eosinophilic oesophagitis: impact of treat- food allergy in pediatric patients with eosi-
eczema: a European multicenter study. ment on desmosomes. J Clin Pathol nophilic esophagitis. Ann Allergy Asthma
Allergy 2004;59:1318–1325. 2015;68:51–56. Immunol 2010;104:496–502.
48. Yamazaki K, Murray JA, Arora AS, 59. Katzka DA, Tadi R, Smyrk TC, Katarya 71. Corridoni D, Arseneau KO, Cominelli F.
Alexander JA, Smyrk TC, Butterfield JH E, Sharma A, Geno DM et al. Effects of Inflammatory bowel disease. Immunol Lett
et al. Allergen-specific in vitro cytokine pro- topical steroids on tight junction proteins 2014;161:231–235.

Allergy 71 (2016) 611–620 © 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 619
EoE: a non-IgE-mediated hypersensitivity Simon et al.

72. Xavier RJ, Podolsky DK. Unravelling the 84. Liacouras CA, Spergel JM, Ruchelli E, 93. Miehlke S, Alpan O, Schroeder S, Strau-
pathogenesis of inflammatory bowel dis- Verma R, Mascarenhas M, Semeao E et al. mann A. Induction of eosinophilic
ease. Nature 2007;448:427–434. Eosinophilic esophagitis: a 10-year experi- esophagitis by sublingual pollen
73. Huttenhower C, Kostic AD, Xavier RJ. ence in 381 children. Clin Gastroenterol immunotherapy. Case Rep Gastroenterol
Inflammatory bowel disease as a model for Hepatol 2005;3:1198–1206. 2013;7:363–368.
translating the microbiome. Immunity 85. Spergel JM, Brown-Whitehorn TF, Beau- 94. Maggadottir SM, Hill DA, Ruymann K,
2014;40:843–854. soleil JL, Franciosi J, Shuker M, Verma R Brown-Whitehorn TF, Cianferoni A, Shu-
74. Cianferoni A, Spergel JM. Food allergy: et al. 14 years of eosinophilic esophagitis: ker M et al. Resolution of acute IgE-
review, classification and diagnosis. Allergol clinical features and prognosis. J Pediatr mediated allergy with development of eosi-
Int 2009;58:457–466. Gastroenterol Nutr 2009;48:30–36. nophilic esophagitis triggered by the same
75. Nowak-Wegrzyn A, Sampson HA. Adverse 86. Arias A, Gonzalez-Cervera J, Tenias JM, food. J Allergy Clin Immunol
reactions to foods. Med Clin North Am Lucendo AJ. Efficacy of dietary interven- 2014;133:1487–1489.
2006;90:97–127. tions for inducing histologic remission in 95. Niranjan R, Mavi P, Rayapudi M, Dynda
76. Simons FE. Anaphylaxis. J Allergy Clin patients with eosinophilic esophagitis: a sys- S, Mishra A. Pathogenic role of mast cells
Immunol 2010;125:161–181. tematic review and meta-analysis. Gastroen- in experimental eosinophilic esophagitis.
77. Nowak-Wegrzyn A, Muraro A. Food terology 2014;146:1639–1648. Am J Physiol Gastrointest Liver Physiol
protein-induced enterocolitis syndrome. 87. Lee D, Albenberg L, Compher C, Baldas- 2013;304:1087–1094.
Curr Opin Allergy Clin Immunol sano R, Piccoli D, Lewis JD et al. Diet in 96. Noti M, Wojno ED, Kim BS, Siracusa
2009;9:371–377. the pathogenesis and treatment of inflam- MC, Giacomin PR, Nair MG et al.
78. Venter C, Groetch M. Nutritional manage- matory bowel diseases. Gastroenterology Thymic stromal lymphopoietin-elicited
ment of food protein-induced enterocolitis 2015;148:1087–1106. basophil responses promote
syndrome. Curr Opin Allergy Clin Immunol 88. Kagalwalla AF, Shah A, Li BU, Sentongo eosinophilic esophagitis. Nat Med
2014;14:255–262. TA, Ritz S, Manuel-Rubio M et al. Identi- 2013;19:1005–1013.
79. Ruffner MA, Ruymann K, Barni S, Cian- fication of specific foods responsible for 97. Schoepfer AM, Safroneeva E, Bussmann C,
feroni A, Brown-Whitehorn T, Spergel JM. inflammation in children with eosinophilic Kuchen T, Portmann S, Simon HU et al.
Food protein-induced enterocolitis syn- esophagitis successfully treated with empiric Delay in diagnosis of eosinophilic esophagi-
drome: insights from review of a large elimination diet. J Pediatr Gastroenterol tis increases risk for stricture formation in
referral population. J Allergy Clin Immunol Nutr 2011;53:145–149. a time-dependent manner. Gastroenterology
Pract 2013;1:343–349. 89. Molina-Infante J, Arias A, Barrio J, 2013;145:1230–1236.
80. Berin MC. Immunopathophysiology of Rodrıguez-Sanchez J, Sanchez-Cazalilla M, 98. Dellon ES, Kim HP, Sperry SL, Rybnicek
food protein-induced enterocolitis syn- Lucendo AJ. Four-food group elimination DA, Woosley JT, Shaheen NJ. A phenotypic
drome. J. Allergy Clin Immunol diet for adult eosinophilic esophagitis: A analysis shows that eosinophilic esophagitis
2015;135:1108–1113. prospective multicenter study. J Allergy is a progressive fibrostenotic disease. Gas-
81. Rocha R, Vitor AB, Trindade E, Lima R, Clin Immunol 2014;134:1093–1099. trointest Endosc 2014;79:577–585.
Tavares M, Lopes J et al. Omalizumab in 90. van Rhijn BD, Vlieg-Boerstra BJ, Versteeg 99. Straumann A. Treatment of eosinophilic
the treatment of eosinophilic esophagitis SA, Akkerdaas JH, van Ree R, Terreehorst esophagitis: diet, drugs, or dilation? Gas-
and food allergy. Eur J Pediatr I et al. Evaluation of allergen-microarray- troenterology 2012;142:1409–1411.
2011;170:1471–1474. guided dietary intervention as treatment of 100. Lucendo AJ, Arias A, Gonzalez-Cervera J,
82. Kagalwalla AF, Sentongo TA, Ritz S, Hess eosinophilic esophagitis. J Allergy Clin Yague-Compadre JL, Guagnozzi D, Angu-
T, Nelson SP, Emerick KM et al. Effect of Immunol 2015;136:1095–1097. eira T et al. Empiric 6-food elimination diet
six-food elimination diet on clinical and 91. Lucendo AJ, Arias A, Tenias JM. Relation induced and maintained prolonged remis-
histologic outcomes in eosinophilic between eosinophilic esophagitis and oral sion in patients with adult eosinophilic
esophagitis. Clin Gastroenterol Hepatol immunotherapy for food allergy: a system- esophagitis: a prospective study in the food
2006;4:1097–1102. atic review with meta-analysis. Ann Allergy cause of the disease. J Allergy Clin Immunol
83. Gonsalves N, Yang GY, Doerfler B, Ritz Asthma Immunol 2014;113:624–629. 2013;131:797–804.
S, Ditto AM, Hirano I. Elimination diet 92. Ridolo E, De Angelis GL, Dall’aglio P. 101. Wen T, Stucke EM, Grotjan TM, Kemme
effectively treats eosinophilic esophagitis in Eosinophilic esophagitis after specific KA, Abonia JP, Putnam PE et al. Molecu-
adults; food reintroduction identifies causa- oral tolerance induction for egg protein. lar diagnosis of eosinophilic esophagitis by
tive factors. Gastroenterology 2012;142: Ann Allergy Asthma Immunol 2011;106: gene expression profiling. Gastroenterology
1451–1459. 73–74. 2013;145:1289–1299.

620 Allergy 71 (2016) 611–620 © 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd