Most of the reported cases of non-EoE-EGID with duodenal ulcers were of male teenagers [10,11,12,13,14,15], and half had a history of atopic complications, such as asthma or allergic rhinitis, similar to our case. A previous study reported that 7.5% of pediatric gastrointestinal ulcers were non-EoE-EGIDs [16]. Non-EoE-EGID cases with duodenal ulcers are often refractory, steroid-resistant, and recurrent [11, 12, 14].

Food is a causative trigger of EGIDs; however, identifying food antigens is often difficult because EGIDs are mainly caused by non-IgE-dependent food allergies. Antigen-specific LST reveals the involvement of non-IgE-independent cell-mediated immune response, and the antigen-specific LST was reported to be positive in a case report of non-EoE-EGID due to chicken eggs [17], as in our patient. However, we were unable to prove that the positive LST result was the cause of non-EoE EGID because atopic patients retain a variety of allergen-specific T-cell responses. Egg patch testing was not performed on this patient. Moreover, LSTs for other foods such as milk, wheat, and soy were not performed on this patient. Although reports on the efficacy of food elimination therapy exist in patients with non-EoE-EGID [18, 19], well-designed high-quality studies are lacking [19]. Our patient showed inadequate remission on an egg-free diet alone and required appropriate pharmacological treatment. Considering that recurrence was observed even after starting the egg-free diet, it was suggested that multiple unknown causative substances may be involved. Systemic steroids are often administered to induce remission in patients with non-EoE-EGID. These steroids can improve symptoms in a short period and have a high remission rate; however, relapses often occur after discontinuation. In a previous long-term cohort study, 95% of patients treated with oral corticosteroids had remission, but 58% of patients subsequently had multiple relapses and a chronic course [4]. Therefore, long-term administration is often necessary and can lead to side effects such as osteoporosis, susceptibility to infections, and glucose intolerance. Considering that our patient had a history of steroid-induced glaucoma, it was necessary to avoid long-term high-dose steroid administration. Careful steroid dose reduction was attempted; however, it could not be discontinued completely because of persistent gastrointestinal symptoms. A retrospective study of pediatric patients with non-EoE-EGID reported that topical budesonide therapy was as effective as systemic steroid administration [20]. Given that our patient's eosinophilic inflammation occurred in the stomach and duodenum, we speculated that oral administration of an inhaled budesonide suspension, which is approved for bronchial asthma as a topical steroid therapy, would be pharmacokinetically effective. We did not select it to treat our patient, because the patient and his parents did not fully agree with the efficacy of topical steroid treatment and strongly desired an early remission of severe atopic dermatitis. Therefore, dupilumab was selected as a treatment for atopic dermatitis after taking informed consent.

The pathogenesis of EGIDs, which are non-IgE-mediated food allergies, involves type 2 inflammation caused by antigen-specific Th2 lymphocytes and antigen-nonspecific innate lymphocyte type 2 (ILC2) [21, 22]. Causative agents such as food that enter the lamina propria of the intestinal tract are captured by the dendritic cells and are presented to the CD4-positive naive T cells. Cell-mediated immunity is induced by persistent stimulation, differentiation of naive T cells to type 2 helper T cells is promoted through IL-4 production, and cytokines, such as IL-4, IL-5, and IL-13, are produced by Th2 lymphocytes. Additionally, innate immune mediators such as thymic stromal lymphopoietin (TSLP) are produced by intestinal mucosal epithelium damage, and IL-5 and IL-13 are produced by ILC2. IL-5 is involved in eosinophil differentiation, survival, and activation, while IL-13 acts on the intestinal epithelium and vascular endothelium to promote eotaxin-3 production, resulting in eosinophil migration and accumulation in the local mucosa. IL-13 also induces periostin expression, promotes fibrosis and smooth muscle proliferation, and is involved in tissue remodeling. Transcriptome analysis of gastric and duodenal mucosal tissues from patients with non-EoE-EGID revealed an increased expression of cytokines involved in Th2-type immune responses, such as IL-4, IL-5, IL-13, and eotaxin-3 [23, 24]. Additionally, increased serum IL-5, IL-13, TARC, TSLP, and eotaxin-3 levels have been reported in patients with non-EoE-EGID [25, 26]. Therefore, the efficacy of various molecularly targeted drugs against Th2 cytokines in EGID is currently under investigation [27]. Mepolizumab is a humanized monoclonal antibody directed against interleukin-5 and is approved for the treatment of multiple type 2 inflammatory diseases, including atopic dermatitis, asthma, chronic rhinosinusitis with nasal polyps, and eosinophilic esophagitis. Two case reports showed that all three patients with non-EoE EGID who received mepolizumab administration showed clinical improvement [28, 29], with one having histologically decreased eosinophil counts. One patient showed increased eosinophil count in the mucosal tissue after mepolizumab treatment. Although mepolizumab may improve clinical symptoms in non-EoE EGID, its efficacy in suppressing eosinophilic infiltration in tissues is not yet fully clear. Future clinical trials on more patients with non-EoE EGID are needed. Dupilumab, an IL-4 receptor alpha monoclonal antibody that inhibits both IL-4 and IL-13 signaling, has been approved for the treatment of type 2 inflammatory diseases such as severe asthma and atopic dermatitis. In a clinical trial in patients with EoE aged ≥ 12 years, dupilumab significantly increased the rate of histologic remission and improved dysphagia after 24 weeks [8]; however, its efficacy in non-EoE EGID has only been reported in sporadic cases [30, 31] (Table 2). All four reported patients were children, three were males, and all four had severe atopic dermatitis or asthma. Elimination of the causative food in three cases, local steroid therapy in three cases, PPIs in three cases, and systemic steroids in two cases did not lead to remission. Dupilumab administration improved endoscopic and histologic findings after 6 weeks to 6 months, allowing one patient to discontinue systemic steroids and another to develop tolerance to the causative food. Similar to these cases, our case showed improvement in endoscopic and histologic findings 5 months after administration; systemic steroid administration was discontinued, and egg tolerance was acquired. Administration of dupilumab may lead to tolerance to the causative foods by suppressing eosinophilic inflammation in the intestine; however, this mechanism is not yet fully understood.

We also investigated the changes in the serum levels of IL-5, IL-13, and eotaxin-3 during treatment. Although no sufficient reduction of eosinophil counts was observed after the initiation of oral prednisolone, normalization of eosinophil counts was observed after the initiation of dupilumab. Similarly, serum levels of total IgE, TARC, IL-5, IL-13, and eotaxin-3 were decreased after administration of dupilumab, as expected in an atopic patient. IL-4R/IL-13R-mediated signaling suppression by dupilumab may have decreased eotaxin-3 secretion from the intestinal epithelium or vascular endothelium, as well as decreased IL-5 and IL-13 secretion from Th2 lymphocytes. Interestingly, the serum levels of IL-5 and eotaxin-3 remained within normal ranges, although the eosinophil counts in the blood increased again after discontinuation of prednisolone. A transient increase in blood eosinophil counts after the initiation of dupilumab has been reported in various atopic diseases such as bronchial asthma [32]. Given the sustained suppression of eosinophil infiltration in the gastrointestinal mucosa in our patient, dupilumab may have suppressed the extravascular migration of eosinophils into the gastrointestinal tract. However, as this is a single case, future large longitudinal cohort studies are required to validate dupilumab efficacy. A randomized clinical trial including the clinical efficacy of dupilumab and biomarker analysis in patients with non-EoE EGID is currently underway (ClinicalTrials.gov number NCT03678545), and the results are highly anticipated.

Additionally, there are no clear criteria for discontinuing dupilumab in patients with EGIDs. A reduction in esophageal eosinophil infiltration and improvements in dysphagia, weight loss, growth failure, esophageal stricture, and pathological fibrosis may be the timing of consideration for discontinuing dupilumab in patients with EoE. Similarly, for non-EoE EGID, reduction in mucosal eosinophil infiltration and improvement in abdominal pain, weight loss, and failure to thrive may be the criteria for dupilumab discontinuation. It is unclear whether long-term administration of dupilumab will lead to tolerance without EGID recurrence or side effects. This may be investigated in future research studies.

In conclusion, we presented a pediatric case of non-EoE-EGID with a duodenal ulcer. Suppression of Th2-type immune responses by dupilumab may improve abdominal symptoms as well as endoscopic and histologic findings, lead to discontinuation of systemic steroid administration, and achieve tolerance to causative foods.