INTRODUCTION — Eosinophilic esophagitis (EoE) was defined by a panel of experts as "a chronic, immune/antigen-mediated, esophageal disease characterized clinically by symptoms related to esophageal dysfunction and histologically by eosinophil-predominant inflammation" [1]. Allergy testing in EoE is warranted to identify and manage common comorbid atopic diseases (eg, asthma, allergic rhinitis) and to determine foods that may present a risk for acute allergic reactions/immediate-type allergy when eliminated foods are reintroduced into the diet during treatment. It may also help identify EoE triggers.
The evidence supporting the role of allergies in EoE and the various types of allergy testing used are reviewed here. Other aspects of disease pathogenesis, clinical manifestations, general diagnostic evaluation, and treatment (including choosing a specific elimination diet) are reviewed separately. (See "Clinical manifestations and diagnosis of eosinophilic esophagitis (EoE)" and "Treatment of eosinophilic esophagitis (EoE)" and "Dietary management of eosinophilic esophagitis" and "Patient education: Eosinophilic esophagitis (The Basics)".)
EVIDENCE FOR ROLE OF ALLERGIES IN EoE — Several lines of evidence suggest that allergies play a significant role in EoE. An immune response to allergen/antigen is a primary factor in the pathogenesis of the disease, and there is evidence that both environmental allergies and food allergies are involved. In addition, the majority of patients with EoE are atopic, with concomitant asthma, allergic rhinitis, atopic dermatitis, and/or immunoglobulin E (IgE) mediated food allergy.
Mechanistic studies — Evidence for allergies in EoE is supported by murine models that show that skin sensitization protocols can induce eosinophilic inflammation in the esophagus when that organ is challenged [2-4]. In addition, examination of esophageal samples has indicated that classic allergic T2 cytokines and chemokines (interleukin [IL] 4, IL-5, IL-13, eotaxin-3) are involved in the pathogenesis of EoE [1,4-7]. Additional details on the pathogenesis of EoE are discussed separately. (See "Eosinophilic esophagitis (EoE): Genetics and immunopathogenesis", section on 'Overview of pathogenesis'.)
Clinical response to dietary changes — The evidence for the role of food allergies in the pathogenesis of EoE is based upon the clinical response when patients are placed on various elimination diets, with resolution of symptoms and normalization of esophageal biopsies. The data are equally extensive and convincing in children and adults. This evidence is discussed in detail separately. (See "Dietary management of eosinophilic esophagitis", section on 'Role of food allergy in EoE'.)
Association with environmental allergies — The evidence that environmental allergies are involved in EoE is more controversial than food allergies, and the precise mechanisms involved are unknown. Environmental allergies may contribute to EoE in allergic patients through the direct effect of swallowed inhaled allergens in the esophagus and/or through postnasal drainage from allergic rhinitis triggering symptomatic and histologic changes [8]. Alternatively, it may involve a regional airway response or a more systemic response with subsequent trafficking of eosinophils into the esophagus [9,10].
The diagnosis of EoE has been shown to occur in a seasonal pattern in both children and adults, with peaks of new diagnosis occurring at the height of the aeroallergen season and a decrease in newly diagnosed patients in the winter [11-13]. In one of these series, less intense esophageal eosinophilia was seen in the winter [13]. Consistent with this result, duodenal eosinophilia, mucosal IgE+ cells, and T cells were found on biopsies during the peak of birch pollen season, but not out of season, in birch-sensitive patients [14]. Up to 20 eosinophils per high power field were found in the esophagus during pollen season in a set of patients with seasonal allergies [15]. A review of patients with EoE at the author's institution found that approximately 14 percent of patients had symptoms that were worse during the pollen season [8]. Histologic changes during the peak pollen season were confirmed in 20 percent of that cohort.
One patient was reported whose EoE was solely dependent on environmental allergies, with up to 40 eosinophils per high power field in the esophagus that resolved out of pollen season [16]. We have found five more similar patients that have no food sensitization but only aeroallergen sensitization [17]. These patients tend to be older than the patients with food-dependent EoE. Consistent with this finding and similar to what is seen in children without EoE, food sensitization based upon skin testing and atopy patch testing (APT) decreased with age and aeroallergen sensitization increased with age in a cross-section of children with EoE [18].
Atopic history — Multiple studies have found that 30 to 50 percent of pediatric patients with EoE have clinician-diagnosed asthma, up to 20 percent have atopic dermatitis, 50 to 75 percent have allergic rhinitis, and 60 percent have urticarial reactions or anaphylaxis to foods [19-25]. There are less data in adults, but the population appears similarly atopic, with 30 to 50 percent having asthma, 50 to 75 percent having allergic rhinitis, approximately 51 percent having food pollen allergy syndrome, and 5 to 10 percent having atopic dermatitis [26-31].
FOOD ALLERGY TESTING IN EoE — Testing for IgE-mediated food allergies in patients with EoE is primarily used to identify patients who are also at risk for acute allergic reactions to foods (eg, urticaria, anaphylaxis) when the foods are reintroduced after an empiric elimination diet [32]. The clinical utility of food allergen testing for identification of EoE triggers remains largely unknown, with data concerning sensitivity and specificity derived thus far from a few centers that primarily manage children. Limited data in adolescents and adults suggest that the value of food allergy testing in EoE appears to decline with increasing age. In addition, allergy testing for foods is not solely sufficient to diagnose food triggers (response to diet and results from biopsies are needed to confirm the relationship). (See "Dietary management of eosinophilic esophagitis", section on 'Reintroduction of foods'.)
There are three main methods for testing food allergies in EoE: skin prick testing (SPT), in vitro specific IgE testing, and atopy patch testing (APT). SPT and in vitro testing are used to identify IgE-mediated allergies, and APT is used to identify non-IgE-mediated allergies.
A food that tests negative is probably not a trigger if the negative predictive value (NPV) for that food-specific test is high. Similarly, a food that tests positive is most likely a trigger if the positive predictive value (PPV) for that food-specific test is high. Predictive values are affected by disease prevalence. APT appears to have better NPVs than SPT (ie, the foods that test negative are probably not triggers), but PPVs are typically better with SPT than APT (ie, the foods that test positive may be triggers). Combining APT and SPT results improves NPV but worsens PPV. (See 'Our approach' below.)
Detailed discussions of skin and blood testing for allergic disease are reviewed separately. Other testing methods including basophil activation testing, lymphocyte stimulation, intradermal skin testing, and others have not been studied in EoE. (See "Overview of skin testing for IgE-mediated allergic disease" and "Overview of in vitro allergy tests" and "Diagnostic evaluation of IgE-mediated food allergy" and "Future diagnostic tools for food allergy".)
Skin prick testing — The predictive values of SPT to individual foods associated with resolution of symptoms and normalization of esophageal biopsies have been explored in patients with EoE. The PPVs and NPVs for 10 of the most common foods that cause EoE were determined in a pediatric population evaluated at one center [33-35]. The NPV on isolated skin testing was generally high, with a rate of greater than 90 percent for less common foods (potato, chicken, peanut, beef, pork, and corn) and 79 to 90 percent for egg, wheat, and soy. However, the NPV was less than 30 percent for milk. The PPVs were poor, ranging from 26 to 62 percent, with milk being the exception at 86 percent.
In a series of 20 adults who had success with the six-food (empiric) elimination diet, food triggers were identified in all patients upon reintroduction [36]. However, 13 of these patients were negative on skin testing to all foods tested, and skin testing only predicted 13 percent of causal foods in this series. In another series, six adults had sensitization to wheat that was not clinically relevant [37]. In a third series, 15 adults underwent a six-week elimination diet based upon SPT, prick-prick testing (PPT) to fresh foods, and APT [38]. Only four patients (26 percent) achieved complete remission, and one patient had clinical remission with partial histologic improvement.
Specific IgE testing — There are few studies that examine the predictive values of specific IgE testing via in vitro immunoassay in EoE. For milk-specific IgE, most studies have not found it to be predictive. One study of 53 patients with EoE found that specific IgE was a more sensitive marker of sensitization compared with APT and SPT, but a major limitation of this study was that it looked at sensitization, not clinically relevant allergy [39,40]. Another small study found that patients with EoE and very low milk IgE (0.10 to 1.00 int. unit/mL) were more likely to respond to milk elimination than those with negative milk IgE (<0.10 int. unit/mL) or higher milk IgE (>1.00 int. unit/mL) [41]. Studies examining component testing have shown no value in this tool for identification of foods causing EoE [42].
Specific IgG4 testing — The exact role of immunoglobulin G4 (IgG4), if any, in the pathogenesis of EoE and its utility in testing is not clear. Data have suggested that IgG4 is a marker of exposure, is not predictive of reaction, and, in patients with IgE-mediated allergy, is a marker for resolution of allergy. Elevated specific IgG4 to typical causal foods are found in adults and children with EoE [43-45]. One study found that detectable levels of food-specific IgG4 (≥0.10 microg/mL) were common in children regardless of whether they had EoE or IgE food sensitization [45]. However, high IgG 4 titers (≥10.0 microg/mL) to several cow's milk (CM) proteins (Bos d 4, 5, and 8) and gluten were more commonly seen in children with EoE. This was particularly true in children with EoE and detectable IgE to CM compared with controls regardless of sensitization status to CM. Specific IgG4 levels decrease by half on average after six to eight weeks on a CM-free diet. (See "Food allergy in children: Prevalence, natural history, and monitoring for resolution", section on 'Mechanisms' and "Unproven and disproven tests for food allergy", section on 'IgG food tests'.)
Atopy patch testing — The use of APT was initially explored for the diagnosis of food allergies in patients with atopic dermatitis and subsequently studied in patients with gastrointestinal symptoms. A general discussion of APT and its use in atopic dermatitis are reviewed separately. (See "Future diagnostic tools for food allergy", section on 'Atopy patch testing'.)
APT involves the application of fresh food or rehydrated food in an occlusive Finn chamber for 48 hours on a patient's back. The Finn chamber is then removed, and a reading of the reactions is done immediately after removal and 24 hours later. Reactions are read using a consensus scale from the European Task Force on Atopic Dermatitis that is based upon the findings of induration, erythema, and papules [46]. This scale is different than the scale used in contact dermatitis, which only examines induration. The main limitation in APT is the lack of standardized reagents. In addition, the studies of APT were primarily performed in patients with known disease, and, therefore, the test may not have the same predictive values for identifying food allergies in the general public. Additional issues include operator dependence, need to keep the allergens on the skin for 48 hours, and the requirement for multiple visits to have the tests placed, removed, and read.
The predictive values for APT in EoE were determined for causal foods identified after normalization of biopsies following removal of single foods or reintroduction of a food leading to an increase in eosinophils (>15 per high power field) on esophageal biopsies performed at one center [33-35,47]. The predictive values for the most common causal foods were calculated in a sequential fashion, with each patient first undergoing SPT and then APT. Patients did not have APT to foods that were positive on SPT. The NPVs for APT were better than for SPT, with values >90 percent for egg, soy, peanut, beef, corn, chicken, potato, and pork and 87 percent for wheat. However, the NPV remained low for milk at 31 percent. The PPVs were poor for most foods (12 to 51 percent), except for milk at 86 percent and egg at 61 percent in a cohort of 319 children [33,35]. Another group found poor NPVs for APT in a cohort of 23 children and young adults who followed testing-directed elimination diets [48]. However, the two groups used different testing materials and reading scales for a positive APT value (one designed for use in food allergy and the other for contact dermatitis), making it difficult to compare the study results. A lower rate of positive APT in older children compared with younger children suggests a limited value in older children and adults [18].
Skin prick and atopy patch testing — The NPVs for a combination of SPT and APT were slightly better than either test alone, but the PPVs were modestly worse in one larger series [35]. NPVs were >96 percent for peanut, beef, corn, chicken, potato, and pork; 93 percent for egg and soy; 88 percent for wheat; and 44 percent for milk. PPVs ranged from 17 to 42 percent for most foods, with the exception of egg and milk (61 and 82 percent, respectively).
ENVIRONMENTAL ALLERGY TESTING IN EoE — Testing for aeroallergen sensitization is frequently warranted to address comorbid asthma and allergic rhinitis, to identify potential aeroallergen triggers of EoE (ie, seasonal exacerbations), and to consider timing of diagnostic biopsies in relation to pollen seasons. Evaluation for environmental allergies by performing standard allergy testing is indicated in all patients that have signs and/or symptoms of allergic rhinitis based upon history and physical examination. (See 'Association with environmental allergies' above.)
OUR APPROACH — In our practice, we discuss two basic approaches to treatment of EoE. One approach is to use medication and the other is to use diet. For diet, we discuss three options: (1) elemental diet, (2) specific (empiric) food elimination, or (3) allergy testing-directed elimination. The success rate for each approach varies from ≥90 percent success for elemental diet to 50 to 70 percent for specific food elimination and 50 percent for allergy testing-directed elimination. For allergy testing-directed elimination, skin prick testing (SPT) and rarely also atopy patch testing (APT) to foods are used to guide dietary avoidance and reintroduction. (See "Dietary management of eosinophilic esophagitis".)
In children and adults with a diagnosis of EoE, our approach to allergy testing and treatment is twofold:
●We perform standard environmental allergy testing for patients that have signs and/or symptoms of allergic rhinitis based upon history and physical examination. Seasonal allergies are treated with standard allergic rhinoconjunctivitis medications. We also treat patients with environmental allergies with intranasal glucocorticoids for one month prior to endoscopy and biopsy because observational data suggest that postnasal drainage from allergic rhinitis may contribute to esophageal eosinophilia [8]. Treatment with nasal steroids does not appear to directly affect esophageal eosinophilia and therefore should not mask active EoE due to food allergies. (See "Allergic rhinitis: Clinical manifestations, epidemiology, and diagnosis" and "Pharmacotherapy of allergic rhinitis".)
●We offer the patient/caregivers two options to determine the food(s) causing EoE:
•Empiric food elimination of the top four allergens (milk, egg, soy, and wheat). Most patients at the author's institution choose this approach instead of the testing-directed approach because of similar rates of efficacy and fewer clinician visits with empiric elimination. For these patients, we perform SPT of the eliminated foods prior to reintroduction.
•Allergy testing with elimination of positive foods (milk is eliminated even if negative on testing). We perform SPT and rarely also APT for the initial food allergy evaluation in the few patients with EoE who choose the testing-directed approach.
For specific (empiric) food elimination, we find that eliminating just milk, egg, wheat, and soy leads to improvement in the biopsy in approximately 50 percent of patients based upon several studies and our own data [49-51]. We have chosen four-food elimination instead of six-food group elimination since peanut, tree nuts, fish, and shellfish are very rare causes of EoE. Testing can be performed or additional allergens eliminated empirically if this approach fails and the patient/caregivers still wish to pursue dietary therapy. Other centers prefer the six-food group elimination diet, which includes peanut/tree nuts and fish/shellfish, since this is the most extensively studied empiric elimination diet. However, there are almost no published cases of EoE triggered by fish, shellfish, or tree nuts and only a few cases of EoE triggered by peanut. Another alternative is elimination of milk, egg, wheat, and all legumes, rather than just soy. (See "Dietary management of eosinophilic esophagitis", section on 'Efficacy of different dietary approaches' and "Dietary management of eosinophilic esophagitis", section on 'Choosing an elimination diet'.)
For those who choose a testing-directed approach for dietary elimination, we first SPT to the top 10 foods (milk, egg, wheat, soy, peanut, beef, corn, chicken, potato, and pork) and the other major foods in the patient's diet (limiting to no more than 15 foods in total). We rarely also perform APT to the foods that were negative on SPT. We remove all positive foods from the diet (testing-directed elimination diet) and also milk since both SPT and APT have poor negative predictive values (NPVs) for identifying milk as a causal agent in EoE. (See 'Skin prick testing' above and 'Atopy patch testing' above.)
A complete discussion of elimination diets, including when to perform repeat endoscopies and food reintroduction, and other treatment options are reviewed in detail separately. (See "Dietary management of eosinophilic esophagitis" and "Treatment of eosinophilic esophagitis (EoE)".)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Eosinophilic esophagitis".)
SUMMARY AND RECOMMENDATIONS
●Eosinophilic esophagitis (EoE) is a chronic, immune-mediated inflammatory disorder of the esophagus. Most patients with EoE have underlying atopic conditions, and food allergies play a major role in the pathogenesis of EoE. (See 'Introduction' above and 'Evidence for role of allergies in EoE' above.)
●The primary method of food allergy testing used for patients with EoE is skin prick testing (SPT). Atopy patch testing (APT) is used infrequently. In general, these tests have good negative predictive values (NPVs) and poor positive predictive values (PPVs) for foods, with the exception of milk. (See 'Food allergy testing in EoE' above.)
●Allergy testing in EoE is used to identify and manage common comorbid atopic diseases (eg, asthma, allergic rhinitis) and to determine foods that may present a risk for acute allergic reactions/immediate-type allergy when eliminated foods are reintroduced into the diet during treatment. It may also help identify EoE triggers. (See 'Food allergy testing in EoE' above and 'Our approach' above.)
