Benralizumab Completely Depletes Gastrointestinal Tissue Eosinophils and Improves Symptoms in Eosinophilic Gastrointestinal Disease

Abstract

BACKGROUND:

Previous studies of targeted eosinophil biologics in eosinophilic esophagitis have yielded mixed results. Possible explanations include incomplete eosinophil depletion with anticytokine (anti–IL-5) treatments and/or irreversible fibrotic tissue changes contributing to symptomatology.

OBJECTIVE:

To characterize the therapeutic effect of eosinophil depletion in patients with hypereosinophilic syndrome with varied eosinophilic gastrointestinal (GI) disorders.

METHODS:

Hematologic, histologic, endoscopic, and clinical symptoms for a subset (n = 7) of hypereosinophilic syndrome patients with GI tissue eosinophilia enrolled in a phase 2 clinical trial of benralizumab (anti–IL-5RA) were assessed before and after treatment (NCT02130882).

RESULTS:

Blood and GI tissue eosinophils were completely depleted in all segments of the GI tract, and all patients reported improved GI symptoms, in some cases as early as after the first monthly dose. Some patients had recurrent symptomatic flares without recurrent peripheral or tissue eosinophilia, in most cases after prolonged symptomatic remission and in the setting of liberalization of dietary restrictions and/or tapering of background therapy. Although eosinophil-associated histologic changes improved in all segments, epithelial changes persisted in the esophagus and stomach in patients with recurrent disease flares even after 1 year of treatment. Serum tryptase and GI mast cells were generally unchanged with treatment, and increases were not associated with disease flares. Serum levels of IL-4 and IL-5 increased with benralizumab treatment (both P < .05).

CONCLUSIONS:

Benralizumab treatment completely depleted blood and GI tissue eosinophilia in patients with eosinophilic GI disorders, but clinical response, while encouraging, was heterogeneous. Residual symptoms in some patients may reflect persistent epithelial changes in the upper GI tract.

INTRODUCTION

Eosinophilic gastrointestinal disease (EGID) is defined by the presence of gastrointestinal (GI) symptoms accompanied by excess tissue eosinophils in single or multiple parts of the GI tract and can present with or without significant blood eosinophilia.^1,2 The most common GI segment involved is the esophagus (eosinophilic esophagitis [EoE]) followed by the stomach (eosinophilic gastritis [EG]), small bowel (when with gastric involvement, eosinophilic gastroenteritis [EGE], or by itself, eosinophilic enteritis [EE]), and colon (eosinophilic colitis). Unlike EoE, EG, EGE, EE, and eosinophilic colitis are often accompanied by blood eosinophilia³ that can rise to the level of hypereosinophilia (blood eosinophilia ≥1500 cells/μL).^4,5 Current therapeutic options for non-EoE EGIDs include topical and systemic corticosteroids and/or empiric food elimination diets. These treatments are associated with side effects, affect quality of life, and have variable efficacy. Disease remission is typically defined by a combination of endoscopic evaluation with reduction in GI tissue eosinophils and symptomatic improvement.

After promising early results,⁶ randomized placebo-controlled clinical trials using IL-5 antibodies (ie, mepolizumab and reslizumab) to treat patients with EoE have been disappointing, with persistence of clinical symptoms despite decreased GI tissue eosinophilia by approximately 50% in those who were treated in most studies.^7–9 One suggested explanation was that persistent symptoms, such as dysphagia in EoE, may represent fibrosis or structural changes that would not be impacted by removal of inflammatory cells, such as eosinophils. An alternate proposed explanation was that symptoms reflected incomplete eosinophil tissue depletion.

Benralizumab is an afucosylated antibody directed against the alpha chain of the IL-5 cytokine receptor (IL-5RA), which is highly expressed on eosinophils and their precursors, and to a lesser degree on basophils and mast cells. Benralizumab is Food and Drug Administration–approved as add-on treatment for moderate to severe eosinophilic asthma and dramatically depletes both peripheral blood and tissue eosinophils. A phase 2 clinical trial of benralizumab (30 mg subcutaneously monthly) in PDGFRA-negative hypereosinophilic syndrome (HES) demonstrated a complete hematologic response in 17 of 19 subjects with undetectable absolute eosinophil counts (AECs) in most cases.¹⁰ Seven of the 20 subjects in the study had HES with GI symptoms and documented GI eosinophilia that met criteria for EGID. All 7 experienced symptomatic improvement on benralizumab, and repeat endoscopic evaluation with tissue biopsy after 3 to 6 months of benralizumab therapy revealed complete depletion of tissue eosinophils in all segments of the GI tract. This study aimed to describe the hematologic, histologic, and symptom findings after 1 year of benralizumab therapy in these 7 patients.

METHODS

Twenty symptomatic, treatment-refractory PDGFRA-negative patients with HES and an AEC of greater than or equal to 1000 cells/μL despite treatment were recruited and enrolled on a clinical trial of benralizumab (NCT02130882¹⁰). The clinical trial consisted of 12 weeks of placebo versus benralizumab 30 mg subcutaneously given every 4 weeks, followed by open-label benralizumab 30 mg subcutaneously given every 4 weeks for up to 48 weeks. Responders were eligible to continue benralizumab on an open-label extension. All study participants were coenrolled on a natural history protocol to study patients with eosinophilia (NCT00001406). Both studies were approved by the National Institute of Allergy and Infectious Diseases Institutional Review Board at the National Institutes of Health. All subjects signed informed consent.

Seven of the 20 subjects enrolled on the original trial (subjects #6, #7, #11, #13, #14, #15, and #16) met criteria for EGID, defined as GI symptoms and historical evidence of GI tissue eosinophilia (≥15 eosinophils/HPF) in esophageal biopsies for EoE, greater than or equal to 30 eosinophils/HPF in stomach/duodenum biopsies for EG/EGE, and more than 60 eosinophils/HPF in colon biopsies for eosinophilic colitis. Because this was not a trial focused on eosino-philic GI disease but on HES in general, patients were not block randomized for GI disease, which resulted in uneven numbers randomized to placebo (n = 2) or drug (n = 5) during the first section of the trial. Endoscopy and tissue biopsies were performed at baseline and at week 24 (after 3 or 6 doses of the drug). All 7 subjects had upper endoscopies with biopsies of every GI segment. Five of the 7 subjects also underwent lower GI endoscopic evalua-tion (colonoscopy), with tissue biopsy of every GI segment in 4 subjects and random biopsy in 1 subject. Three patients had an additional endoscopy with biopsies performed after 1 year. Biopsies were reviewed and scored by National Institutes of Health pathol-ogists using both standard of care and research protocols (M.H. Collins, MD, unpublished data). Mast cells were enumerated in 5 HPF using tryptase immuno-staining (clone AA1; Dako). GI tissue lymphocytes were qualita-tively evaluated using CD3 (Clone 2GV6; Roche), CD4 (Clone SP35; Roche), CD8 (Clone SP57; Roche), and CD20 (clone L26; Roche) immunostaining using an automated staining system (Benchmark ULTRA, Roche) according to the manufacturer instructions.

Serum total IgE and tryptase levels were measured as part of the parent clinical trial. Serum cytokines (IL-4, IL-5, IL-13, IFN-γ, and IL-17A) were evaluated by multiplex analysis (Millipore) following manufacturer’s instructions. Statistical analyses of mast cell counts and serum cytokine, tryptase, and total IgE levels were performed using nonparametric matched-pairs signed rank test (Wilcoxon).

RESULTS

The clinical characteristics of the 7 subjects with HES with significant GI symptoms and biopsy-proven GI tissue eosino-philia are provided in Table I. The subjects ranged in age from 23 to 53 years (median age, 35 years) and were predominantly (5 of 7) female. Six of the 7 subjects had more than 1 affected GI segment, and symptoms tended to correspond with the segments demonstrating GI eosinophilia on biopsy. Most (5 of 7) subjects were on medical and/or dietary therapies for their disease at clinical trial entry and had previously failed multiple therapies, including other experimental therapies. Despite treatment, these patients continued to experience debilitating GI symptoms (indicated in bold in Table I) and had significant blood eosinophilia with AECs exceeding 1000 cells/μL, meeting the inclusion criteria for the trial. Comorbid atopic disorders were common: 3 of 7 had significant IgE-mediated food allergies and 5 of 7 carried a diagnosis of asthma.

TABLE I.

Baseline clinical characteristics of patients

	Age (y), sex		GI tissue eosinophilia*
No.		GI symptoms (active, historic)	E	G	D	C	Comorbid atopic disease	Therapies (active, historic)
6†	48 F	Diarrhea (3× daily) Abdominal pain Dysphagia	−	+	−	+	Asthma	P, HU, IFN
7	35 M	Bloating Constipation Abdominal pain (daily, with monthly exacerbation) Diarrhea, weight loss, dysphagia, odynophagia, nausea, early satiety	+	+	+	−	Asthma, FA (crab, almond; avoids shellfish and nuts), allergic rhinitis	P, topical budesonide, FV, hydroxychloroquine, sirolimus‡
11	53 M	Abdominal pain Diarrhea (3× /wk) dysphagia early satiety, weight loss	+	+	+	−	Asthma	P, topical budesonide
13	37 F	Diarrhea (5–7× daily) Nausea (daily) Pill dysphagia Abdominal pain	−	+	+	+		P, Elemental diet, enterocort, cromolyn
14	23 F	Abdominal pain/cramping (daily) Diarrhea (multiple times a day) Dysphagia	+	+	+	−	Asthma, FA (wheat, suspected dairy)	Elemental diet, dietary therapy, topical budesonide, cromolyn, FV, sirolimus‡
15†	35 F	Diarrhea Abdominal pain Intermittent dysphagia	−	−	−	+		P, FV, IFN
16	33 F	Nausea (constant, weekly worsening) Dysphagia (daily)	+	+	−	−	Asthma, AR, FA (shrimp)	P, IFN, FV, cromolyn, topical budesonide, sirolimus,‡ dietary therapy

Budesonide, Budesonide slurry; C, colon; D, duodenum; E, esophageal; F, female; FA, food allergy; FV, swallowed flovent (fluticasone); G, gastric; HU, hydroxyurea; P, prednisone.

Historic symptoms and therapies as well as active GI symptoms and therapies at trial enrollment (bolded) are listed.

^*Tissue eosinophilia meeting criteria as described in methods.

^†Subjects with lymphoid HES whose disease also included GI manifestations. Remaining subjects had HES single-organ overlap (EGID).

^‡Sirolimus given under phase I clinical trial.

Five of the 7 subjects with HES with eosinophilic GI manifestations were randomized to drug for the first 12 weeks, and 2 were randomized to placebo; the latter 2 subjects received benralizumab as part of the open-label protocol after 12 weeks. Peripheral blood eosinophilia dramatically decreased in all 7 subjects within 1 day of receiving the active drug (Figure 1, A) and remained suppressed throughout the remainder of the 48-week trial (Figure 1, B) with 1 exception (subject #6).

FIGURE 1.

AECs in the 7 subjects with eosinophilic GI disease in (A) the week after receiving their first dose of benralizumab and (B) over the 48 weeks of the trial. For those randomized to drug (solid symbols), Figure 1, A, depicts the first week of the clinical trial. For those randomized to placebo (open symbols), Figure 1, A, depicts the first week in the open-label phase (week 12).

All 7 subjects reported clinical improvement, but the timing and degree of improvement varied by individual. Some subjects reported improved GI symptoms as early as after 1 or 2 doses of the active drug, whereas others did not report improvement until after 4 or 5 doses (see Table E1 in this article’s Online Repository at www.jaci-inpractice.org). By week 48, 4 of 7 subjects had either tapered medical therapy and/or expanded their diet or altered their food consumption patterns (Table II). Although subject 6 reported clinical improvement and was able to taper off hydroxyurea by 1 week after her first dose and slowly reduce her prednisone to 5 mg, she unfortunately relapsed at week 32 with return of eosinophilia and symptoms (Figure 1, B). By week 48, she had resumed prednisone 15 mg daily (Table II). As previously reported,¹⁰ the loss of efficacy in this subject was not due to antidrug antibodies.

TABLE II.

Benralizumab allowed some subjects to reduce medical therapies for EGID and liberalize dietary restrictions

No.	BL Tx	Diet at BL	Week 48 Tx	Diet at week 48
6*	P 10 mg daily, HU 1 g daily	Unrestricted	P 15 mg	Unrestricted
7	Budesonide 9 mg FV 440 twice a day	Avoids: chicken and beef plus FA foods	Budesonide 6 mg, FV 440 twice a day	No attempt to reintroduce foods; eats more freely
11	Budesonide 3 mg	Unrestricted	None	Unrestricted
13	None	Avoids: carrots, peas, apple, mango, banana, watermelon, kiwi, strawberry, turkey, direct egg, corn, wheat, oats	None	Reintroduced most foods Avoids: wheat and mango
14	FV 220 2p twice a day	Consumes only shrimp, potato, fruit (apple, pear, lemon, banana), vegetables (broccoli, lettuce)	FV 220 2p twice a day	Reintroduced fish, poultry, tomato, carrots. Unable to tolerate avocado and shrimp
15*	None	Unrestricted, but planned mealtimes to avoid symptoms	None	Eats unrestricted diet freely
16	FV 220 2p twice a day	Avoids: corn, rice, most meats, shrimp, shellfish, and pure egg	FV 220 2p twice a day	No attempts to reintroduce foods

Budesonide, Budesonide slurry; P, prednisone; FA, food allergy; FV, swallowed fluticasone; HU, hydroxyurea; Tx, treatment.

^*Subjects with lymphoid HES and GI involvement. Remaining subjects were HES single-organ overlap (EGID). Subject #6 relapsed at week 32 and taken off benralizumab.

Endoscopic evaluation and biopsies were performed at baseline and at week 24 (after receipt of 3 or 6 monthly doses of benralizumab). Endoscopic findings in the esophagus and stomach (n = 7) were mildly improved at week 24 as compared with baseline (see Table E2 in this article’s Online Repository at www.jaci-inpractice.org). The duodenum (n = 6) and colon (n = 5) had normal endoscopic appearances in all subjects studied at baseline and at week 24. On histological examination, tissue eosinophils were completely depleted in all GI segments examined at week 24 (Figure 2¹⁰). Eosinophil granule protein staining at week 24 confirmed the absence of tissue eosinophils in all cases examined.¹⁰

FIGURE 2.

Histological features (hematoxylin and eosin) of gastric biopsies at baseline and week 24 at 200×. (A) Subject 16 at baseline with increased eosinophils in the lamina propria and glandular epithelium, and epithelial changes with marked (>5 HPF) increased nuclear:cytoplasmic ratio. (B) Subject 16 at week 24 with persistent epithelial changes consisting of marked (>5 HPF) increased nuclear:cytoplasmic ratio and reduced mucin despite resolution of tissue eosinophils. (C) Subject 16 at 1 year with persistent epithelial changes. (D) Subject 14 at baseline with increased eosinophils in the lamina propria and glandular epithelium, and mild/moderate epithelial changes (≤5 HPF) consisting of reduced mucin and increased nuclear:cytoplasmic ratio. (E) Subject 14 at week 24 with persistent epithelial changes with reduced mucin and increased nuclear:cytoplasmic ratio found in more than 5 HPF despite resolution of tissue eosinophils. Black arrows, eosinophils; blue arrows, areas of increased nuclear:cytoplasmic ratio and reduced mucin; inset (black box, 400×).

The effect of eosinophil depletion on concomitant medical/dietary therapies or food avoidance patterns was heterogeneous (Table II). At baseline, 2 subjects (#7 and #15) carefully arranged their mealtimes to avoid having GI symptoms during critical parts of their work day, but with eosinophil depletion, they were able to eat more freely. Two other subjects (#13 and #14) were able to reintroduce specific foods back into their diet. One subject (#14) became intolerant of a food (shrimp) that they had tolerated at the beginning of the trial. Four subjects (#7, #11, #14, and #16) reported partial return of symptoms or episodic worsening of symptoms at approximately 1 year, typically in the context of tapering off medications or food reintroduction. One subject continued to have episodic flares of disease, albeit improved over baseline, without change in diet or medication (#16). Peripheral blood eosinophil count remained suppressed (AEC = 0) at 1 year in all 6 subjects with EGID who remained on benralizumab therapy.

Endoscopic evaluation was repeated at 1 year in 3 subjects (#7, #11, and #16). In all 3 cases, tissue eosinophils remained undetectable (Figure 2, C).

To determine whether continued symptoms were the result of other effector cells, additional immunohistochemical evaluation of biopsy specimens was performed. Tryptase staining demonstrated that mast cell numbers (reported as the average number in ≥5 HPF) were not significantly different at baseline and 24 weeks in the group overall (Figure 3, A). Individual subjects demonstrated either an increase or a decrease in GI tissue mast cell counts, but the direction did not correlate with the presence of symptoms. Moreover, serum tryptase levels (a measure of total mast cell burden) were measured at baseline and every 3 months thereafter and were unchanged throughout the course of the trial (Figure 3, B). Immunohistochemical stains for T and B lymphocytes in GI tissue demonstrated the presence of both cell types at baseline, with CD3⁺ T cells being predominant. There was no qualitative change after treatment with benralizumab (data not shown).

FIGURE 3.

(A) MC numbers in tissue at baseline and after 3 to 6 doses of benralizumab treatment. Tryptase-positive MCs were enumerated in an average of 5 HPF per sample. (B) Serum tryptase levels over the 48 weeks for each subject. Benra, Benralizumab; BL, baseline; MC, mast cell.

Baseline, week 24, and week 52 tissue biopsies were reviewed in detail using both published and unpublished scoring mechanisms that highlight changes related to both eosinophils and epithelial cells (Collins et al¹¹ and M.H. Collins, MD, unpublished data). In biopsies from subjects who had recurrent flares, tissue epithelial features such as spongiosis and basal zone hy-perplasia improved in grade and stage but persisted in the esophagus. Furthermore, reactive epithelial changes in the stomach persisted or worsened despite depletion of eosinophils (Figure 2). Interestingly, all subjects with primary colonic involvement experienced disease remission without recurrent flares, with colonic mucosa depleted of eosinophils but otherwise normal in appearance at week 24 (Figure 4).

FIGURE 4.

Histological features (hematoxylin and eosin) of colonic biopsies at baseline and week 24 at 200×. (A) Subject 13 at baseline with increased eosinophils in the lamina propria and glandular epithelium. (B) Subject 13 at week 24 with resolution of tissue eosinophilia.

Murine and limited human studies suggest that eosinophils play an immunoregulatory role in maintaining homeostasis in the GI tract through their effects on T_H2 cells (reviewed in Klion et al¹²). To determine whether eosinophil depletion resulted in disruption of this putative immunoregulatory role leading to symptom recurrence in patients with GI involvement treated with benralizumab, serum T_H2 cytokines (IL-4, IL-5, IL-13) as well as IFN-γ and IL-17A were measured at baseline and 48 weeks. Serum IL-4 and IL-5 levels, but not IL-13, IFN-γ, or IL-17A, were significantly increased at 48 weeks (geometric mean IL-4: 2.95 vs 35.61 pg/mL, P = .047, and IL-5: 6.38 vs 110.5 pg/mL, P = .0156, paired Wilcoxon rank; see Figure E3, B, in this article’s Online Repository at www.jaci-inpractice.org). Total IgE levels were also examined and, in all but 1 subject (#14), remained stable throughout the course of the trial (Figure E3, A).

DISCUSSION

In patients with HES with GI involvement, benralizumab depleted blood eosinophils within 1 day and GI tissue eosinophils at the first time point examined (24 weeks). Although eosinophil depletion persisted through 1 year of treatment, with associated improvement in symptoms in all patients, the degree of symptomatic improvement and the time it took for this to occur were variable. Potential explanations for this variability include (1) a lag between eosinophil depletion and normalization of the mucosa and (2) involvement of other cells, such as epithelial cells or mast cells, in the pathogenesis of symptoms.

Published studies in adults and children with EoE have described persistence of basal zone hyperplasia and spongiosis despite reduction of tissue eosinophilia and improvement of symptoms with either dietary or corticosteroid treatment.^13,14 In the present study, associated epithelial changes were demonstrated in both affected esophageal and stomach biopsies despite undetectable tissue eosinophils, even after 1 year of treatment. More importantly, these findings were accompanied by persistence of symptoms. The finding that epithelial changes persist in the stomach despite complete eosinophil depletion has not been previously reported. In contrast, there were few to absent epithelial changes associated with colonic eosinophilia in the biopsies of symptomatic subjects at baseline, similar to what has been recently reported in larger studies,¹⁵ and the colonic histology normalized with eosinophil depletion, paralleling the complete clinical response experienced by the subjects. These data suggest important differences between the pathogenesis of EGIDs in the different GI segments.

There is increasing interest in understanding the role of mast cells in EGIDs. In our study, GI tissue mast cells were enumerated. Their levels were unaffected by anti–IL-5RA (benralizumab) treatment and were similar between subjects who did and did not have persistent symptoms and histological changes. Consistent with these findings, serum tryptase levels were unchanged over the course of the study. This lack of evidence for mast cell involvement differs from published data in pediatric EoE, where increased tissue mast cells were associated with abnormal endoscopic findings consistent with EoE and continued symptoms despite histologically inactive disease, defined as tissue eosinophils less than 15 cells/HPF,¹⁶ but is consistent with reported data in the stomach and duodenum. In a recent study of patients comparing patients with EG/EGE (n = 52) to patients without EGID who underwent endoscopy and biopsy for upper GI symptoms (n = 123), mast cells were measured in the stomach and duodenum. Peak mast cell numbers were significantly increased in the EG/EGE cohort in the stomach (44.2 ± 19.6 compared with 21.9 ± 8.2 cells/HPF) and duodenum (47.5 ±14.8 compared with 28.1 ± 9.3 cells/HPF).¹⁷ Although mast cell numbers were correlated with eosinophil number in the tissues of patients with EG/EGE, there was no correlation between GI symptoms and higher mast cell counts in either cohort, suggesting that mast cell counts alone are not determinative of symptoms.¹⁷ A recent phase 2 study of the anti–Siglec-8 mAb lirentelimab, which depletes eosinophils and reduces mast cell activation,^18,19 demonstrated tissue eosinophil depletion and improvement of symptoms in patients with EG; symptom scores were reduced by more than 30% in only 27of 43 patients.²⁰ Whether these 27 patients had higher levels of tissue mast cells or increased mast cell activation was not addressed.

Finally, serum cytokine data in the present study provided little evidence for a major effect of benralizumab and eosinophil depletion on the global cytokine milieu. Of the cytokines measured, only IL-4 and IL-5 were significantly changed over the course of the study. Increases in serum IL-5 have been previously reported in benralizumab trials in asthma²¹ and may reflect the loss of the main consumer of this cytokine, the eosinophil. An increase in serum IL-4 may explain the increases in total IgE in some patients, especially subject #14, because it is a driver of B-cell isotype switching. The significance of this observation is yet to be determined, however, because it was not universal among the subjects with recurrent or persistent symptoms.

Dietary therapies through specific food avoidance are one of the mainstays of therapies for EGIDs of the upper tract (eg, EoE and EG) but they carry significant burden and impact on quality of life. One hope of both clinicians and patients is that newer therapies could allow reintroduction of specific foods previously avoided. If the eosinophil is the sole or primary effector cell in EGID, one might expect that complete elimination of this cell with benralizumab should allow one to reintroduce specific foods previously avoided to reduce eosinophilic inflammation. Larger, carefully designed studies and validated biomarkers for food triggers will be needed to address this mechanistically and clinically important question. Finally, the eosinophil is suggested to be important in metabolism, host defense, and malignancy. Although this study is small, the lack of deleterious effects noted in our hypereosinophilic subjects depleted of eosinophils suggests that this cell type may either play a limited role in the above-mentioned processes and/or alternate compensatory pathways exist.

Limitations of the present study include the small number of subjects and the lack of validated symptom scores in subjects with GI symptoms, because the original study was not designed to focus on eosinophilic GI disease and symptom scoring tools were not available when the phase 2 clinical study was initiated. The strengths of the current retrospective study include the length of follow-up and the breadth as well as detailed analyses of individual subject clinical and laboratory data.

In summary, our data contribute to the emerging body of data that suggest that there may be differences in presentation, pathogenesis, and treatment responses depending on the segment of GI involvement in EGIDs.^3,22,23 Notably, despite the small numbers, subjects with primary colonic involvement report substantial improvement in symptoms after benralizumab treatment. In aggregate, most subjects experienced improvement in symptoms, providing further evidence that eosinophils are important drivers of disease pathogenesis in EGID.

CONCLUSIONS

Targeted eosinophil treatment with benralizumab depleted tissue eosinophils below the level of detection in patients with HES with GI tissue involvement. Although eosinophil depletion was associated with significant symptomatic improvement, diversity in clinical and histologic responses between patients over time and in different GI segments suggests that factors other than eosinophils likely play a role in disease pathogenesis in at least some settings. Clinical trials in larger cohorts of patients with EGID with single and multiple GI segment involvement using therapies that target eosinophils alone (NCT04543409, NCT03473977, NCT03656380) or in conjunction with other cell types (NCT04322604) are needed to address these issues.

What is already known about this topic?

Reduction or normalization of gastrointestinal (GI) tissue eosinophilia with diet or corticosteroid therapy is associated with clinical remission in patients with eosinophilic GI disease.

What does this article add to our knowledge?

Although benralizumab dramatically depletes GI tissue eosinophils and improves symptoms in treatment-refractory patients with hypereosinophilic syndrome and GI eosinophilia, clinical response over time is heterogeneous and may depend on the GI segment affected.

How does this study impact current management guidelines?

Benralizumab may be useful for the treatment of patients with hypereosinophilic syndrome and GI eosinophilia that is refractory to standard therapy. Larger studies are needed to confirm these findings.

Abbreviations used

AEC

absolute eosinophil count

EG

eosinophilic gastritis

EGE

eosinophilic gastroenteritis

EGID

eosinophilic gastrointestinal disease

EoE

eosinophilic esophagitis

HES

hypereosinophilic syndrome