Eosinophilic Esophagitis Pathogenesis: All Clear? Open Access

Eosinophilic esophagitis (EoE) is a food- and aeroallergen-driven, type 2 chronic inflammation characterized by >15 eosinophils in a genetically predisposed individual triggered by environmental factors (shown in Fig. 1 [1]). Increasing numbers of publications since the first description of EoE by Atwood and Straumann in 1993 [2, 3] have implicated the rise in EoE incidence and prevalence due to the exposure to detergents [4], changes in the microbiota composition [5, 6], impaired epithelial integrity [7], and type 2 inflammation [8] with many of the past findings being associative. The field moves with sophisticated studies and the implementation of meticulous clinical trials [9] from associations toward translational/mechanistic advances in the pathophysiology of EoE. This movement, escorted by the development of novel technologies, such as organoid cell culture [10] and high-throughput sequencing technologies [11], will push the field forward to understand the complex mucosal immune system of the esophagus and the development of EoE, resulting in novel therapies for EoE.

The esophageal mucosal immune system can be pharmacologically manipulated to treat EoE. A recent example is the introduction of the monoclonal antibody dupilumab into the clinic, which targets the alpha chain shared by the IL-4 and IL-13 receptors [9]. Further advances in EoE research are essential in light of the possible debilitating disease course with food impaction [12], perforation [13], and impaired life quality [14] for patients suffering from EoE. Genome-wide association studies (GWAS) have linked the EoE risk with single-nucleotide polymorphisms (SNPs) in genes implicated in barrier integrity and type 2 immunity [15]. The rise of the EoE prevalence in the last decades cannot be attributed to genetic predisposition alone. Studies have described associations between EoE and exposure to pollutants impairing barrier integrity, such as food emulsifiers in industrialized food, laundry, and toothpaste detergents [16]. Providing causal relationships between the exposome and EoE development is challenging, but we anticipate it will be critical to offer causal relationships to improve the care of EoE patients.

Recent years have witnessed tremendous progress in molecular biology and mucosal immunology technologies, including metagenomics for microbiome sequencing, T-cell receptor (TCR) sequencing, three-dimensional cell culture technologies, and metabolomics. In parallel, clinical trials have significantly improved by defining more stringent endpoints. These advancements will help uncover new approaches for treating EoE by restoring the disturbed epithelial barrier, targeting immune pathways, manipulating the esophageal microbiota, preventing exposure to environmental triggers, and better understanding how food-derived antigens induce EoE.

We want to emphasize that the esophageal immune system and microbiome are understudied compared to the skin and the small and large bowel [17]. We predict that the increased incidence and prevalence of EoE, with significantly impaired quality of life for affected individuals, and the opportunities for the pharmaceutical industry to develop biologics for EoE will arouse the interest of academic and pharmaceutical industry-driven research in the esophageal immune system. In this review, we aimed to summarize the current knowledge of EoE pathophysiology and discuss open questions in EoE research.

The incidence and prevalence of EoE are rising, as studies with children and adults in North and South America, Europe, Asia, and Australia suggest [18‒22]. It is estimated that the EoE prevalence is 32.5 in adults and 30.9 in children per 100,000 population, affecting 1 in 3,000 inhabitants in areas with a westernized lifestyle [1]. The steady increase in EoE, previously considered a rare disease, raises the question of why this dramatic increase in prevalence occurs. This increase in EoE prevalence coincides with an increase in allergies, autoimmune diseases, and inflammatory bowel disease [23]. The hygiene hypothesis proposes that reduced exposure to certain microorganisms and parasites in early childhood, as observed in westernized areas, predisposes to increased frequencies of allergies and autoimmune diseases [24]. In light of the hygiene hypothesis, an inverse correlation between EoE and gastric Helicobacter pylori infection has been reported [25‒28]. It must be pointed out that EoE, in contrast to allergies, develops independently of IgE and depends on food and aeroallergens [29]. More importantly, epidemiological studies investigating the associations between children’s exposure to microorganisms and an increased EoE risk provide conflicting results. A case-controlled study investigating associations between perinatal factors, such as preterm delivery, neonatal intensive care unit (NICU) admission, breastfeeding, antibiotics in infancy, and absence of furry pets and EoE development revealed associations between breastfeeding and early NICU admission in children with SNP in calpain-14 (CAPN14), CCL26, thymic stromal lymphopoietin (TSLP), the KLF13 region, and TGFB [30].

Conversely, Radano and colleagues [31] found associations between caesarean sections, early antibiotic use, and EoE. Studies reporting associations between living in urban and rural areas with EoE provided inconsistent results. In a cross-sectional, case-control study of patients with esophageal biopsies in a US national pathology database, living in rural areas was associated with increased esophageal eosinophilia [30]. Another cross-sectional study with children <18 years enrolled in the US Medicaid reported that living in rural areas and areas with more neighborhood poverty had lower odd risk associations with EoE [32]. Possible confounding factors by race and ethnicity and different definitions of urban areas may have contributed to divergent results, as in the USA, more Caucasians with a higher risk of EoE live in rural areas. More detailed studies delineating environmental factors associated with EoE development are warranted to explain the increased EoE prevalence in westernized areas.

Another theory has hypothesized that an increased barrier leakiness due to pollutants is implicated in the rising prevalence of allergies, autoimmune diseases, and inflammatory bowel disease [33]. This hypothesis seems particularly relevant for EoE as the penetration of food and aeroallergens across the epithelial barrier causes EoE. Plasticizers, such as bisphenol A and phthalate added to polymers or plastics to soften them, have also been suggested to impair the epithelial barrier and be associated with type 2 inflammation [34]. It has been proposed that changes in industrialized food processing occurred parallel with the increase in EoE prevalence. Emulsifiers create stable homogenous emulsions by binding their hydrophobic ends to oils and fats and hydrophilic ends to water. They are food additives in bread, chocolate, ice cream, margarine, and processed meat. Observational studies have indicated associations between ultra-processed food and type 2 inflammation, including EoE [35]. Providing causal relationships between food emulsifiers and EoE is challenging as prospective randomized-controlled trials investigating the impact of food emulsifiers on gastrointestinal diseases must be conducted in the future.

Recently, detergents, such as sodium dodecyl sulfate (SDS), present in laundry detergents and toothpaste and used as a food additive, have been suggested to be associated with EoE. An exciting report shows that adding increasing concentrations of SDS to air-liquid interface (ALI) cultures with the telomerase-immortalized human esophageal epithelial cell line EPC2 and to mouse esophageal organoids impairs esophageal barrier integrity assessed by transepithelial electrical resistance (TEER) measurements [4]. Adding 0.5% SDS to the drinking water for 14 days elicited esophageal inflammation in 50% to 100% of C57BL/6J mice, depending on the cohort, associated with eosinophilic abscesses and spongiosis [4]. The authors used low SDS concentrations in the in vitro (200 ng/mL) and in vivo (0.5% SDS in the drinking water) experiments, which were lower than the 3% SDS in some toothpaste. However, future studies must still address whether comparable concentrations and exposure times occur in humans when exposed to SDS in detergents and hygiene products.

Studying the link between changes in food production and processing, including livestock farming and plant breeding, together with industrial food production that has occurred in parallel with the emergence of EoE, is an exciting avenue in EoE research. Considering the increasing number of observational cohort studies describing associations between changes in the food production of westernized countries, providing evidence of causal relationships between ultra-processed foods and EoE remains challenging. In the future, interventional clinical trials or preclinical experimental systems will establish potential causal connections between environmental pollutants, dietary behaviors, and EoE.

The esophageal mucosal immune system has been implicated as critical for the development of EoE. Since the esophageal immune system is in continuous contact with food-derived antigens and microorganisms that colonize the surface of the epithelial cells, it is tempting to speculate that an altered esophageal microbiome is involved in EoE. A study in mice indicated regional differences in the microbiome composition between the proximal and distal esophagus, reflecting anatomic differences in the muscle composition between the proximal and distal esophagus [36]. The exposure of the distal esophagus to gastric reflux with a low pH may, in addition, also explain the differences in the microbiome composition between different esophagus segments. In the same study, germ-free animals were colonized with a microbiome from SPF animals, inducing transcriptional changes in genes involved in epithelial development, cell adhesion, nuclear body proteins, cellular respiration, and histone acetylation. This suggests that although the density of the esophageal microbiome is low compared to the large intestine, possibly explained by the rapid transit of particles through the esophagus [17], the microbiota influence tissue homeostasis and perhaps epithelial barrier integrity [36]. Challenges in obtaining samples by invasive gastroscopy, amplifying low microbe numbers before sequencing, and preventing oral and gastric cross contamination are technical hurdles to studying the microbiome’s composition in the esophagus. This may explain why only a few studies have investigated associations of the esophageal microbiome with EoE. Active EoE generally leads to an increased microbial density with increased Haemophilus abundance [5, 37] and decreased representatives of the Firmicutes phylum, including Streptococcus, Lactobacillus, Veillonella, and Parvimona [6, 38]. One study reported normalization of the microbiota composition in patients on a six-food elimination diet [6]. Other studies have conversely shown increased microorganism abundance in EoE independent of the treatment and disease activity [5].

Interestingly, the esophageal microbiome correlates with the oral microbiome [6, 39, 40]. Possible explanations for the correlation between the oral and esophageal microbiome could be that oral microorganisms are constantly swallowed into the esophagus. The salivary glands produce IgA that binds to and coats microorganisms through canonical binding with their Fab fragments and other noncanonical binding modalities, including glycan regions on the Fc fragment independent of the B-cell repertoire [41]. The swallowing of IgA from the oral cavity leads to the same IgA molecules in the oral cavity and the esophagus, possibly another reason for the correlations between the oral and esophageal microbiome. These findings have to be validated in further studies, taking into account the esophageal and oral biogeography considering the proximal and distal esophagus and the soft buccal mucosa, the keratinized mucosa of the palate, gingiva, and tongue papilla, the tooth enamel, and the saliva in the oral cavity [42]. If independent studies can confirm the correlation between the oral and esophageal microbiome, these findings could open new avenues for EoE research, given the accessibility of the oral cavity. Elegant studies in ileostomy and colostomy patients providing easy access to the small intestinal microbiome without requiring ileocolonoscopy with bowel preparations have demonstrated the adaptability of the microbiome to test meals and the feasibility of investigating the adaptation of the microbiome to the diet when the microbiome is easily accessible [43]. Assuming the potential correlation of the oral and the esophageal microbiome and the easy accessibility of the oral cavity, it will be feasible to study alterations of the microbiome in response to different diets in the context of EoE by swabbing the oral cavity.

Environmental factors partially explain the increase in EoE incidence and prevalence in the past decades. On the other hand, relatives of EoE patients have a 10- to 64-fold increased risk of developing EoE compared to the general population, implicating the importance of genetics for EoE. In monozygotic twins, the unaffected twin has a risk of 41% of developing EoE; in dizygotic twins, the EoE risk of the unaffected twins is 21%; in siblings, the EoE risk is 2.4% [44]. Nonetheless, genetic heritability appears to be a minor risk factor for EoE compared to the exposome, meaning that genetics unfolds its effects in the context of the environment. GWAS have identified SNPs with relatively low odds ratios ranging from 0.682 to 1.949 [15].

Conversely, rare Mendelian disorders, such as connective tissue disorders, including the Marfan syndrome [45], the Loeys-Dietz syndrome [46], Netherton’s syndrome [47], or Erbin deficiency [48], develop EoE-like inflammation in the esophagus [49]. For example, loss of function variants in the receptor of TGFβ (TGFBR1 and TGFBR2) results in the predisposition of EoE in patients with Loeys-Dietz syndrome. Recently, the M318R mutation predisposing humans to allergic diseases has been introduced into the Tgfbr1 locus of mice. These transgenic Tgfbr1^M318R mice with the introduced point mutation driving a TGFβR1 without kinase activity develop EoE-like inflammation of the esophagus [50], indicating that inherited mutations can lead to EoE in rare circumstances.

More commonly than rare inherited mutations are SNPs associated with EoE with low odds ratios that, together with environmental exposures, predispose to EoE. These GWAS are a starting point for mapping the genetic landscape of EoE. Whole exome sequencing studies have yet to be performed in EoE to describe the genetic landscape in EoE. A recent metanalysis has highlighted SNPs in 13 genes associated with EoE [15]. These EoE candidate risk genes include CAPN14, TSLP, WDR36, BACH2, TNFRSF10B, TNFRSF10C, GATA3, EMSY, LRRC32, ATP10A, CLEC16A, PNPLA3, and SAMM50 implicated for epithelial integrity and differentiation of immune cells. These GWAS thus suggest that the epithelium and immune cell differentiation are essential pathways for developing EoE.

GWAS have identified SNPs in genes responsible for epithelial integrity, suggesting impaired epithelial integrity is disease-inherent in EoE [15]. Conversely, immune cell-derived cytokines, such as IL-13, disturb epithelial integrity in the esophagus (shown in Fig. 2 [51, 52]). Treating EoE with elimination diets, proton-pump inhibitors, topical steroids, or monoclonal antibodies, such as dupilumab, targeting IL-4 and IL-13 signaling by binding to the IL-4Rα subunit shared by the IL-4 and IL-13 receptor could restore the impaired epithelial integrity caused by cytokine signaling [9, 51]. An intact epithelium prevents the intrusion of food antigens, providing a barrier restricting the access of potentially harmful antigens into the esophagus. This is particularly relevant for EoE as antigen deposition and penetration in and through the esophageal epithelium have been implicated in EoE [53]. Tight and adherens junctions seal the epithelium to provide the esophagus with a physical barrier adapted to the challenges of environmental signals caused by ingested food, caustic substances, infections, and refluxed gastric contents.

The esophageal epithelium consists of nonkeratinized stratified squamous cells organized in the superficial stratum corneum, the underlying stratum spinosum, and the stratum germinativum [54]. The cells in the stratum corneum have a flat shape condensed through filaggrin (FLG) that binds to the keratins, forming the cytoskeleton of the cells. GWAS have proposed SNPs in the filaggrin genes as risk genes for EoE. As mentioned above, immune cell-derived cytokines also impair epithelial integrity. The most prominent example is IL-13, which has received much attention following the approval of dupilumab for EoE treatment.

Interestingly, stimulating esophageal epithelial cells with IL-13 leads to transcriptional changes resembling the EoE transcriptome [55]. Elegant mouse work using mouse lines with deletion of the Il13ra1 in esophageal epithelial cells indicated that IL-13, not IL-4, is the critical cytokine that alters the epithelium EoE [51]. IL-13 is also involved in epithelial-mesenchymal transition [56, 57], contributing to fibrosis development. Dupilumab, as a new therapeutic option, might prevent and even reverse fibrosis.

In addition to IL-13, transforming growth factor-β1 (TGF-β1) and oncostatin M have been suggested to be involved in esophageal barrier disturbances observed in EoE. Mediated by oncostatin M, mast cells decreased FLG, desmoglein 1, involucrin, and serine peptidase inhibitor Kazal type (SPINK) 7 expression in the immortalized esophageal keratinocyte cell line EPC2, resulting in reduced TEER and increased epithelial permeability [58]. TGF-β1 also reduced TEER by decreasing the tight junction protein claudin-7 expression [59]. Finally, we recently added IL-19, IL-20, and IL-24 to the cytokines that induce esophageal barrier dysfunction. The IL-20 cytokine family members reduced the expression of genes and proteins involved in epithelial cell differentiation and keratinization, including FLG, in patient-derived organoids. They decreased TEER and increased epithelial permeability in patient-derived ALI cultures [52].

The epithelium also produces cytokines, triggering EoE. TSLP SNPs identified on locus 5q22 are risk gene variants for EoE [60]. Epithelial-derived TSLP activates dendritic cells and basophils to induce type 2 immunity in an antigen-induced EoE model in mice [61]. In addition, IL-33, an alarmin released by epithelial cells, acts on type 2 innate lymphoid cells in the lung or intestine [62]. It still needs to be established if this is the case in EoE. In this context, a 12-year-old boy was described as having developed EoE due to a 0.58-Mb duplication at chromosome 9p24.1 encompassing the entire IL33 gene, associated with elevated IL-33 levels [63]. Transgenic animals that overexpress secreted and active IL-33 by the esophageal epithelium spontaneously develop EoE, further supporting the relevance of IL-33 in type 2 inflammation [64]. These findings emphasize that the epithelium is actively involved in the pathogenesis of EoE by producing alarmins and cytokines that drive type 2 inflammation. The epithelium also maintains its integrity through transcription factor expression, such as hypoxia-inducible factor-1α. Hypoxia-inducible factor-1α maintains the esophageal epithelial barrier in response to inflammation-associated tissue hypoxia. It has been implicated that this process is disturbed in EoE [65].

Recently, it has been shown that a delicate balance between proteases and their inhibitors has to be maintained to secure epithelial integrity in the esophagus. Disturbance of this delicate balance is associated with the development of EoE, achieved by either the upregulation of proteases or the downregulation of their inhibitors. Elevated IL-13 increases the expression of the calcium-activated calpain protease CAPN14, degrading desmosomal proteins and impairing epithelial integrity in EoE [66]. Identifying CAPN14 by GWAS as a risk gene for EoE emphasizes the importance of this system in EoE development [67]. Reducing the SPINK family members’ SPINK5 and SPINK7 expression in EoE results in the overactivation of proteases, including kallikrein 5, also contributing to epithelial barrier impairment [68, 69]. Cytokines, such as IL-13, regulate these fine-tuned relationships between proteases and their inhibitors. They are in parallel disease-inherently dysregulated, as GWAS have identified genes coding for proteases to be associated with EoE [67].

Dissecting disease-inherent mechanisms or consequences of inflammation in EoE is challenging. Animal models can provide insight into how inflammation alters epithelial integrity, which usually neglects the genetic predisposition in humans, as inbred mouse strains have the same genetic background to minimize variation between different experiments caused by various genetic backgrounds. It is also possible to stimulate three-dimensional ALI or organoid cultures derived from human esophageal biopsies or the telomerase-immortalized human esophageal epithelial cell line EPC2 to investigate the impact of immune cell-derived mediators on the intestinal epithelium. Conversely, introducing genetic mutations in transgenic animals allows the study of rare inherited mutations, as exemplified in the Tgfbr1^M318R mouse line harboring a point mutation of the Tgfbr1 locus [50]. Genetic tools will also enable the manipulation of epithelial cell lines or primary cell cultures to investigate the impact of defined genetic alterations on epithelial integrity in EoE.

The identification of pathways that restore the epithelium offers excellent potential for the treatment of EoE. Most current therapies target inflammatory cytokines and do not primarily preserve or restore epithelial integrity. Reports suggest that vitamin D can antagonize the effects of IL-13 on the epithelium [70] and that the short-chain fatty acids butyrate and propionate can reverse the epithelial damage caused by IL-13 [71]. Given the low microbial load in the esophagus compared to the colon, it needs to be established whether butyrate serves as an energy reservoir for the esophageal epithelium as it does in the colon, where the fatty acid oxidation of short-chain fatty acids accounts for 70% to 90% of energy and 70% of oxygen consumption of epithelial cells [72]. Furthermore, 17β-estradiol also protects against IL-13-induced epithelial alterations [73]. However, taking into account the associations of elevated 17β-estradiol and frequencies of endometriosis, leiomyomata, venous thromboembolism, breast cancer, and ovarian cancer, we assume that establishing appropriate 17β-estradiol concentrations in females and males is a significant obstacle to avoid side effects. Considering the low vitamin D concentrations in many regions worldwide, testing the protective effects of vitamin D supplementation on EoE severity in interventional trials could be an exciting approach.

EoE is a chronic type 2 inflammation of the esophagus occurring in genetically predisposed individuals exposed to unknown environmental triggers in childhood and adulthood. EoE is characterized by eosinophils (>15 eosinophils per high power field), other immune cells, including mast cells and T cells, and elevated IL-5 and IL-13 expression in the esophagus. Conversely, only a few immune cells reside in the non-inflamed esophagus, and no eosinophils are present in the esophagus, in contrast to other gastrointestinal tract segments, such as the small and large intestines. Better characterization of the esophageal immune landscape is required to understand the esophageal immune system in physiology and diseases, considering that the mucosal immunology of the esophagus is relatively understudied compared to the small and large intestine [17]. The increased eosinophil numbers in EoE imply that eosinophils are essential for the development of EoE. Clinical trials with eosinophil-depleting antibodies reduced eosinophil numbers in the esophagus as expected. Yet, they did not resolve clinical symptoms, presuming that other immune cells are involved and critical for the pathogenesis of EoE [74]. In contrast to atopic diseases, IgE does not play a significant role in the development of EoE [29]. Instead, elevated serum IgG4 levels have been described in EoE patients [75]. IgG4 has the lowest prevalence of all IgG subclasses, accounting for <5% of total IgG. Repeated allergen exposure leads to elevated serum IgG4 concentrations in patients with allergic diseases [76]. In line, high titers of IgG4 with specificity to cow milk in the serum and esophageal wall IgG4 deposits have been observed in EoE [29, 77]. Despite the unique properties of IgG4 in Fc-dependent effector functions, including complement activation, antibody-dependent cellular toxicity, and the capability of IgG4 to exchange Fab arms to bind two unrelated antigens [78], the significance of elevated IgG4 in EoE remains unclear.

There are only a few immune cells present in the healthy esophagus. A better understanding of pathways involved in homing immune cells to the esophagus promises the possibility of interfering with antibodies or small molecules. Immune-derived cytokines, including IL-13, induce the expression of chemokines, such as eotaxin-3 (CCL26), that attract eosinophils, T cells, and mast cells to the esophagus [79]. Case reports have suggested that blocking the integrin α4β7 required for immune cell homing to mucosal sites with the monoclonal antibody vedolizumab could be an option for treating EoE [80]. The sphingosine-1-phosphate 1 receptor modulator etrasimod inhibiting lymphocyte egress from lymph nodes is currently tested in the phase II VOYAGE trial (NCT04682639) for their safety and efficacy in EoE. In line with these findings, a recent manuscript reports that pathogenic effector Th2 cells in the esophagus with the GPR15 ligand GPR15L are expressed by epithelial cells in pediatric EoE patients. These findings highlight the relevance of pathways involved in lymphocyte homing to the esophagus for EoE [81]. In this study, using improved single-cell RNA-Seq platforms to profile esophageal tissue and the TCR repertoire, the authors succeeded in capturing and sequencing eosinophils that were previously missed due to the release of RNA-degrading ribonucleases by the eosinophils during the isolation process. The authors found indications for the clonal expansion of pathogenic effector Th2 cells in the esophagus that diverged from non-Th2 cells in other tissues. These clonotypes emerged in milk-reactive pathogenic effector Th2 cells in patients with milk-triggered EoE. In the future, chimeric antigen receptor T cells could eliminate the identified clones, assuming the rapidly progressing field of cellular therapies, including T cells carrying engineered chimeric TCRs directed against the T-cell clonotypes in patients with milk antigen-triggered EoE.

Considering that eggs, wheat, soy, peanuts/tree nuts, and fish/shellfish provoke EoE, it is interesting to investigate whether clonotypes generated by the antigen driving the disease occur in these patients. Nutrition influences multiple physiological processes besides T-cell immunity. A recent interventional trial assessing the impact of a ketogenic and vegan diet on immunity and the microbiota revealed that a ketogenic diet is associated with an enrichment of pathways involved in adaptive immunity. Conversely, the vegan diet was associated with enriching pathways involved in innate immunity [82]. The advancement of computational tools, including artificial intelligence, can predict three-dimensional structures of proteins to anticipate possible interactions and functions based on shape and binding sites in food-derived products, such as AlphaFold2 [83]. Biochemical studies will provide experimental validation to confirm the predictions of computational tools. This knowledge and interventional dietary trials on proteins that influence metabolic pathways will be essential in linking effective dietary interventions to EoE, offering insights into disease mechanisms and potential therapeutic targets.

The development of strictures is a devastating clinical consequence of EoE, leading to food impaction, elevated risk of esophageal perforation, and the need for repeated endoscopic dilatations [57]. It is assumed that fibroblasts in the lamina propria beneath the epithelium release extracellular macromolecules, like collagens and glycoproteins, in response to signals derived from immune and epithelial cell, such as TGFβ, TNF, IL-4, and IL-13, to form the extracellular matrix. Eosinophils induce the production of fibronectin and collagen I by human esophageal fibroblasts and esophageal muscle cells in a TGFβ-dependent manner, demonstrated in cell culture experiments of fibroblast and muscle cells stimulated with eosinophil sonicates in the presence of a neutralizing anti-TGFβ antibody [84]. In line, findings of human fibroblast cultures obtained from pediatric EoE patients confirmed the essential function of TGFβ in the activation of fibroblasts [85], and the genetic deletion of the TGFβ signaling mediator SMAD3 attenuated esophageal remodeling in the ovalbumin-induced experimental EoE mouse model [86]. In addition, TGFβ facilitates esophageal smooth muscle cell contraction, highlighting the essential role of TGFβ in esophageal remodeling in EoE [87].

Active EoE elevates multiple cytokines, suggesting that various cytokines have overlapping and redundant functions in promoting esophageal remodeling. Reductionist approaches focusing on a single cytokine help delineate a particular cytokine’s function in esophageal remodeling during EoE. At the same time, it neglects the complexity of biological systems, considering the >60 cytokines involved in mucosal immune responses. Elegant work using patient-derived three-dimensional organoids indicates that TNF and TGFβ act synergistically in inducing lysyl oxidase, a collagen cross-linking enzyme involved in esophageal remodeling [88]. It is important to note that one cytokine can induce signaling through different receptors, and multiple cytokines can signal through the same receptor. Receptor chains are also shared between different cytokine receptors, further contributing to the complexity of studying EoE. An example is the alpha chain of the IL-4 receptor shared between the IL-4 and the IL-13 receptors. The deletion of the IL-4 receptor alpha chain in squamous epithelial cells in mice reduced esophageal wall thickening and collagen deposition in the oxazolone EoE mouse model [51]. Based on these experimental findings in mice, it can be inferred that the recently approved antibody dupilumab binds to the IL-4 receptor alpha chain and will also impact esophageal remodeling in EoE.

In general, it is believed that the progressive fibro-stenotic EoE depends on the disease duration, as a delay in EoE diagnosis increases the risk for stricture formation [89]; a recent mouse study pointed to the possibility that aging predisposes to fibrosis development. Klochkova and colleagues [90] applied an ovalbumin-induced EoE model to young animals (3 months old) and aged mice (18 months old). The authors did not find significant differences in lamina propria and epithelium eosinophil numbers per high power field, indicating that young and aged mice have comparable inflammation degrees. The aged animals exhibited increased lamina propria thickening, suggesting a more progressive fibrosis development could occur in older individuals. The epithelial cells of the aged mice produced more fibroblast-stimulating factors, including basal fibroblast growth factor and osteopontin. This study highlights that in addition to disease duration and chronicity, individual patient-intrinsic factors, such as age, influence fibrosis development.

After considering genetic predisposition, environmental factors, including exposure to detergents, the richness and complexity of the esophageal microbiota, dietary-derived factors, and the complexity of the immune system and epithelial biology, there is a need to standardize experimental systems and choosing the appropriate system for the relevant scientific questions. As Albert Einstein has coined, “Researchers have to make things as simple as possible. But not easier.” Discussions include patient-derived cell cultures, experimental animal models, and interventional trials to offer insights into disease mechanisms and reveal new therapeutic targets through standardizing experimental systems. Genetic modifications allow the study of the impact of one specific gene on disease development, and clinical trials in a well-defined patient cohort provide the possibility to investigate the effect of a defined target on EoE. It is too simplistic not to anticipate that the different genetic backgrounds of individuals will not influence the trial outcome and that different environments, including the microbiotas in different vivaria, do not influence the results achieved in transgenic animal models of the same genetic backgrounds [91]. Several animal models have been proposed for studying EoE, each with specific advantages and disadvantages (shown in Table 1). Generally, experimental EoE can be induced in animals through exposure to detergents, Aspergillus fumigatus, or sensitization to model antigens, including ovalbumin, peanut extracts, or oxazolone, followed by a subsequent challenge with the same antigen [97, 98]. These inducible models mimic human disease by resembling histological features of human EoE, such as eosinophil infiltration in the mucosa and submucosa. It needs to be pointed out that the models with exposure to A. fumigatus [99], ovalbumin [61], or oxazolone sensitization [51] are immune-mediated after a sensitization phase. The applications of detergents, such as SDS, destroy the integrity of the epithelium [4], whereas, in the latter models, the esophageal epithelium stays intact during the disease induction. The ovalbumin and peanut extract models are IgE-mediated, opposite to human EoE, which is considered IgE-independent. The advantages of these models are the defined onset of the disease, the possibility to investigate the disease course after genetic deletion of defined genes in mice, and the opportunity to test potential drug candidates for the treatment of EoE.

Conversely, EoE can spontaneously develop in animals after genetic modifications, as discussed in the Tgfbr1^M318R mouse line [50], by overexpressing IL-33 in the esophageal epithelium [64], or by installing cytokines, such as IL-13, in the airways [100], thus mimicking rare mutations in humans associated with EoE. These mouse lines predict the function of defined genes or their mutations in EoE and help delineate pathways relevant to EoE.

Mouse models most commonly used in immunology research do not fully recapitulate the human immune system, as these organisms had to adapt to different ecological niches during evolution [101]. In contrast to humans, mice have significant bronchus-associated lymphoid tissue, possibly reflecting the higher exposure to airborne antigens in their environmental niche by living on the ground. Moreover, the peripheral blood of humans is enriched in neutrophils in contrast to mice, which have more lymphocytes in the peripheral blood [101]. Advances in primary cell-derived three-dimensional culture models help translate the findings from experimental mouse models into patient-based models and test their relevance for human EoE. In particular, patient-derived organoids and ALI cultures allow studying the effects of immune-derived cytokines, growth factors, and other mediators on the esophageal epithelium. Kasagi and colleagues [10] with human esophageal organoids derived from the immortalized human esophageal squamous epithelium cell line EPC2-hTERT and patient-derived cells together with Zhang et al. [102] using induced pluripotent stem cell-derived esophageal organoids discovered the significance of TGFβ and bone morphogenetic protein (BMP) inhibition for esophageal progenitor cell development and the crucial role of Notch signaling in the differentiation of the stratified squamous epithelium. Organoids are powerful tools for studying tissue architecture and the expression of potential target genes after stimulation with cytokines. Because of their three-dimensional growth with an inner keratin core, TEER or macromolecule flux cannot be measured in organoids, limiting their use in functional studies investigating epithelial barrier integrity.

Conversely, ALI cultures modeling epithelial differentiation are obtained by growing EPC2-hTERT or primary patient-derived esophageal keratinocytes on semipermeable Transwell membranes followed by differentiation and stratification through increased Ca²⁺ concentration (1.8 mm) in the culture medium and subsequent removal of the medium from the upper Transwell chamber [55]. These ALI systems can be used to investigate the effects of cytokines, such as IL-13 or members of the IL-20 cytokine family, on epithelial integrity [7, 52]. Combining patient-derived organoids and ALI cultures to investigate tissue architecture and integrity of the epithelial barrier in EoE will be applied in EoE studies. We expect to witness a rise in publications using patient-derived three-dimensional cell culture to deduce the effects of immune mediators on the esophageal epithelium.

Insights into the pathophysiology of EoE will be gained from experimental models with targeted genetic manipulation and patient-derived cell culture models, reducing the complexity of biological systems and focusing on networks between the immune system and the epithelial compartment. These systems will provide information on single pathways while, at the same time, neglecting the multidimensional framework of EoE encompassing the role of genetics, the environment, dietary factors, and the complexity of the immune system. Clinical trials that test the targeting of specific pathways with monoclonal antibodies will ultimately prove the significance of a pathway for EoE (shown in Table 2). The approval of dupilumab for the treatment of EoE highlights the importance of preclinical translational research that has elucidated the significance of IL-13 in the development of EoE [9]. Clinical trials with negative results also give exciting insights into EoE. The anti-IgE antibody omalizumab did not improve symptoms or esophageal eosinophilia, highlighting that EoE occurs independently of IgE, contrary to typical IgE-mediated food allergies [29]. Interestingly, different clinical trials and clinical case series testing and describing the anti-IL5 antibodies mepolizumab and reslizumab as well as the anti-IL-5Rα antibody benralizumab showed improvements of eosinophilia but did not result in full resolution of clinical symptoms [74, 107, 108]. Considering that >15 eosinophils/hpf are one criterion for EoE diagnosis, the results of these clinical trials open up questions on the function and importance of eosinophils in EoE. Other immune cells, including T cells, are possibly involved in EoE. These studies underscore the need to better comprehend and characterize the functionality of immune and mesenchymal cells in the esophagus, including fibroblasts and the epithelium, to understand their role in EoE.

A recent trial comparing a one-food elimination diet versus a six-food elimination diet for the treatment of EoE reported histological remission rates in 34% in the one-food elimination diet (cow milk) versus 40% in the six-food elimination diet (cow milk, eggs, wheat, soy, peanuts/tree nuts, and fish/shellfish) [109]. The remission rates in this randomized prospective trial are lower than in a recent metanalysis that suggested remission rates ranging from 67.9% with a six-food elimination diet to 56.9% with a four-food elimination diet, and 42.1% with a two-food elimination diet and 54.1% with a one-food elimination diet [110]. However, another trial with children comparing a one-food elimination diet with a four-food elimination diet resulted in comparable remission rates as the latter study that compared the one-food elimination diet with the six-food elimination diet [111]. Maladherence to the diet, differences in the defined endpoints, or the possibility that additional proteins contribute to the disease development is possible explanation for the divergent results in different trials. Elemental diets that remove all potential food allergens from the food have an efficacy of 71%, stressing that host variables and environmental factors also contribute to EoE [112, 113].

We look forward to advancements in TCR sequencing and the identification of protein-binding sites for food-derived molecules on receptors with AlphaFold2, which will provide exciting insights into EoE pathophysiology in the future. The rapidly increasing TCR engineering technologies could provide opportunities to eliminate T-cell clones with high avidity and functionality to food-derived antigens.

This review discusses the established aspects of EoE pathophysiology and addresses the following open points that will expand our understanding of the disease.

• 1.

  Information on environmental factors associated with EoE needs to be more consistent. Some reports describe an association between breastfeeding and NICU admission with EoE, while other studies describe an association between caesarean sections and early antibiotic use with EoE. Conflicting results exist whether living in rural areas is associated with an increased risk of EoE or not. More epidemiological studies are needed to adjust possible confounders to reveal environmental triggers associated with EoE. These possibilities could be tested in experimental EoE models with germ-free animals or mice with low-diversity microbiota.

• 2.

  The exposure to environmental pollutants, such as the detergent SDS present in laundry detergents and toothpaste, has been proposed to be associated with EoE. The recently developed mouse model in which SDS given in drinking water induces EoE [4] can reveal causal relationships suggested by previous epidemiological studies. This mouse model complements previously established inducible experimental models in which the disease occurred after sensitization to allergens and a subsequent challenge. These models will allow us to test relations between risk genes and potential pollutants in EoE pathophysiology.

• 3.

  GWAS are powerful tools to identify risk genes associated with EoE. The biology of the >13 risk genes and the significance of individual SNPs not necessarily located in the coding regions of the identified gene or not necessarily resulting in a loss of function of the protein is challenging. Whole exome sequencing studies have not yet been performed in EoE. We anticipate that whole exome sequencing studies and the generation of mutants with the CRISPR/Cas9 technology will help elucidate the function of the identified risk genes. This will be a critical first step in personalizing the treatment according to the disturbed pathway in individual patients.

• 4.

  The validation of potential correlations between the oral cavity and the esophageal microbiome provides excellent opportunities for future research, given easy accessibility to the oral cavity. Considering that a one- or six-food elimination diet has multiple yet not-defined effects on the microbiome, the metabolome, and the immune system with potential relevance to EoE, this could give insights into the molecular effects of dietary therapies and the reason why they do not work for everyone with EoE.

• 5.

  An impaired epithelial barrier characterizes EoE. GWAS have identified risk genes associated with EoE that are implicated in epithelial barrier integrity, pointing to an underlying barrier defect in EoE. Conversely, immune-mediated cytokines, including IL-13 and IL-20 cytokines, can impair epithelial barrier integrity, suggesting that the impaired barrier is a consequence of inflammation. Strategies to repair an impaired barrier could be the treatment of inflammation with antibodies such as dupilumab or by applying nutritional supplements, for instance, butyrate or vitamin D, as suggested by experimental data.

• 6.

  The esophageal immune system is understudied. Characterizing the immune landscape in the esophagus in healthy individuals and EoE by mass cytometry or single-cell RNA sequencing will help comprehend and appreciate the esophagus’s immune system. Advances in TCR sequencing will also give information on food-derived peptides recognized by the TCR. Progress in TCR engineering could help eliminate T-cell clones involved in EoE development.

• 7.

  Tissue remodeling, fibrosis development, and strictures are devastating consequences of EoE. Limited information is available on pathways involved in the fibrosis development in EoE. More importantly, how tissue remodeling, fibrosis, and strictures can be pharmaceutically resolved is of significant interest to patients with long-standing EoE suffering from complications that demand repetitive dilatation.

Advances in technologies and clinical trials have improved our understanding of EoE and led to the development and approval of therapies, such as orodispersible budesonide formulations and dupilumab. Type 2 chronic inflammation triggered by food and aeroallergens develops in a genetically predisposed individual exposed to environmental factors impacting the esophageal epithelial barrier. Future studies will uncover the multifaceted interplay between environmental factors, the esophageal microbiome, the epithelium, and the immune system.

Figures have been created with BioRender.

None of the authors have a conflict of interest related to this article.

Swiss National Science Foundation (SNSF) Grant 310030_219210 has supported J.H.N. in writing this manuscript.

Both authors, J.H.N. and T.K., jointly wrote the manuscript and designed the figures.