Abstract
Objectives:
Eosinophilic gastritis (EoG) and duodenitis (EoD) are rare conditions that are poorly understood. Our aim was to describe the natural history of children with varying degrees of gastric or duodenal eosinophilia with respect to disease complications and histologic and endoscopic longitudinal trajectories.
Methods:
The electronic medical record at a tertiary children’s hospital was queried to identify patients with EoG, EoD, or EoG+EoD who were cared for between January 2010–2022. Multiple logistic regression was carried out to explore associations between baseline features and persistence/recurrence of eosinophilia or complications remote from diagnosis.
Results:
We identified 151 patients: 92 with EoG, 24 with EoD, 12 with EoG+EoD, and 23 with tissue eosinophilia but did not meet histologic criteria for EoG or EoD (low grade). Average age at diagnosis was 10.6 years and average follow-up was 5.8 years. Twenty five percent of patients with EoG or EoD had persistence/recurrence of eosinophilia; this was associated with increases in eosinophilic gastritis endoscopic reference score (EG-REFS) (aOR 1.34, CI 1.03–1.74) on diagnostic endoscopy. Eighteen percent suffered disease complications, and development of late complications was associated with presenting with a complication (aOR 9.63, CI 1.09–85.20), severity of duodenal endoscopic abnormalities (aOR 8.74, CI 1.67–45.60), and increases in EG-REFS score (aOR 1.70, CI 1.11–2.63).
Conclusions:
Patients with gastric and duodenal eosinophilia should be followed closely to monitor for recurrence and complications, especially those presenting with endoscopic abnormalities or complications.
Keywords: eosinophilic gastritis, eosinophilic duodenitis, natural history, eosinophilic gastrointestinal diseases
Graphical Abstract
Introduction:
Eosinophilic gastroenteritis was first described over 90 years ago, and is broadly defined by the presence of eosinophilic inflammation in the stomach and/or small intestine along with symptoms suggesting gastrointestinal dysfunction and no identified alternate etiology.1,2 It is less common than eosinophilic esophagitis (EoE), affecting 5–8/100,000 individuals,3,4 but can co-occur with EoE and is also associated with atopy and a Th2-inflammatory signature.5–9 Unlike EoE which is chronic, a number of other longitudinal trajectories have been described in eosinophilic gastroenteritis including spontaneous remission or episodic flares.10–12 Patients have distinct severe disease complications including anemia, gastrointestinal hemorrhage, and perforation.11,13,14
Pending consensus nomenclature recommending more precise terms like eosinophilic gastritis (EoG) and duodenitis (EoD) and recent diagnostic guidelines, the term “eosinophilic gastroenteritis” was liberally applied to patients with variable locations and degrees of eosinophilic inflammation.15,16 This made it challenging to characterize these disorders, and the long-term outcomes and disease complications of EoG and EoD remain sparsely described.5,9,15,17 Our aim was to better characterize the clinical course of children with gastric or duodenal eosinophilia. We hypothesized that gross endoscopic findings at diagnosis would positively associate with persistent/recurrent eosinophilia and the development of late complications in EoG and EoD since similar associations have been described in inflammatory bowel disease18–20 and eosinophilic esophagitis21,22. We retrospectively identified a cohort of pediatric patients with gastric and or duodenal eosinophilia and longitudinal follow-up, described their clinical features, and evaluated these associations.
Methods:
Study Design and Identification of Longitudinal Cohort
This retrospective cohort study was performed at Children’s Hospital Colorado. The electronic medical record was queried using SNOMED CT terms to identify patients with gastric or duodenal eosinophilia over a 12 year period (Figure 1b). Pathology reports were reviewed and patients with gastrointestinal symptoms were classified based on eosinophils/0.237mm2 high-powered-field (eos/hpf) as: EoG (≥30 eos/hpf), EoD (≥50eos/hpf), low-grade gastric eosinophilia (<30eos/hpf), or low-grade duodenal eosinophilia (<50eos/hpf).9,16,17 When eosinophils were not enumerated but descriptors like “innumerable,” “sheets,” or “markedly expanded” were present, the histopathology was re-examined by a pathologist (BN or MA) and eosinophils were enumerated. Patients without gastric or duodenal eosinophilia or with a concomitant diagnosis of inflammatory bowel disease (IBD) were excluded. Additionally, those with less than two years of follow-up or those whose diagnostic clinical, procedure, or pathology reports were unavailable were excluded.
Figure 1:
A. Temporal trends in EoG/EoD diagnosis.
B. Flow chart illustrating patient identification for longitudinal analysis. *chronic gastritis n=8, juvenile polyposis syndrome n=2, erosive esophagitis n=2, eosinophilic granulomatosis with polyangiitis n=1, collagenous gastritis n=1, tracheoesophagel fistula n=1, familial mediterranean fever n=1
Demographic features including biologic sex, past medical history, age at diagnosis, and presenting symptoms were recorded. Hemoglobin levels, peripheral eosinophil counts, and albumin levels were recorded at diagnosis and follow-up endoscopies, as was endoscopic information including treatment prior to and after endoscopy and gross endoscopic and histologic features. EG-REFS scores for the body and antrum were retrospectively assigned (composite score 0–31).23 As a measure of endoscopic duodenitis severity, the number of duodenal abnormalities including erythema, erosion, ulceration, villous blunting, or salmon patches were recorded and summed to generate a novel EoD-Endo Score. One point was assigned to each finding to generate a maximum score of 5.
Natural History
Patients were categorized into two groups based on whether gastric or duodenal eosinophilia above the aforementioned thresholds (stomach ≥30eos/hpf; duodenum ≥50eos/hpf) was noted during follow-up: 1) “persistent/recurrent”: those with eosinophilia that persisted despite treatment or recurred after treatment cessation, and 2) “resolved”: those in whom tissue eosinophilia resolved and did not recur. In a preliminary analysis (n=30) the presence or absence of symptoms attributable to EoG/EoD at time of follow up endoscopy was not recorded in 40%; longitudinal trajectories were therefore based primarily on histologic activity. The primary outcome was the proportion of patients with persistent/recurrent gastroduodenal eosinophilia. We age- and sex-matched twenty EoG patients with persistent/recurrent eosinophilia to patients with resolution, then assessed and compared baseline EoG histologic scores.5
Late Complications
The secondary outcome was the proportion of patients with complications after diagnosis. The presence and timing of disease complications including upper gastrointestinal bleeding (UGIB), moderate or severe anemia without UGIB (hemoglobin <9.9mg/dL if <59 months old, <10.9mg/dL if >59 months old),23 hypoalbuminemia (albumin <3.5mg/dL), bowel perforation, bowel obstruction, surgery, and hospitalization were collected. Complications were categorized as occurring early (prior to/at diagnosis) or late (after diagnosis).
Temporal Diagnostic Trends
The number of patients diagnosed with EoG or EoD annually irrespective of follow-up duration was recorded, as was the number of patients undergoing upper endoscopy annually.
Analytic approach
Parametric data were summarized with means and standard deviations, and between group differences were compared using T-tests and ANOVA. Nonparametric data were described using medians and ranges with between group differences compared using Kruskal-Wallis tests. Categorical data were reported as percentages and compared using Fisher’s exact and Chi2 tests. Least squares regression was used to characterize temporal trends in diagnosis. Univariate logistic regression was used to determine the relationship between baseline features (age, sex, segment involved, complication at diagnosis, peripheral and tissue eosinophil counts, EG-REFS, endoscopic duodenitis severity) and patients with/without persistent/recurrent gastroduodenal eosinophilia or late complications, then significantly (α 0.05) associated features were included in multivariate logistic regression models.
Results:
Temporal Diagnostic Trends
There were 252 patients diagnosed with EoG or EoD during the 12-year study period. The number of patients diagnosed annually trended upward during the study period (Figure 1a). This trend was less pronounced after adjusting for annual upper endoscopy volume.
Clinical Characteristics of Longitudinal Cohort
There were 151 patients with abnormal gastric and/or duodenal eosinophilia and over two years of follow-up: 128 patients with EoG or EoD and 23 patients with low-grade gastric or duodenal eosinophilia (Figure 1b). At diagnosis 92 patients had EoG, 24 patients had EoD, and 12 patients had both (EoG+EoD) (Table 1). The majority of the patients with EoG or EoD had a personal history of atopy (n=91, 71%). The most common atopic comorbidities were allergic rhinitis (n=52, 41%) and food allergy (n=49, 38%). Twenty four percent (n=31) had previously undergone endoscopic evaluation and the majority of these (n=23, 74%) were diagnosed with EoE. Average age at diagnosis for EoG or EoD was 10.6 years old and the most common presenting symptom was abdominal pain (n=69, 54%) (Table 1). At diagnosis, the 18% (n=23) of patients with EoE were on dietary or pharmacologic treatment and an additional 41% (n=52) were pretreated with proton pump inhibitors (PPI). There were no significant differences in any of the aforementioned characteristics between patients with EoG, EoD, or EoG+EoD (Table 1).
Table 1:
Clinical characteristics at the time of EoG/EoD diagnosis
| EoG, EoD, or EoG+EoD n=128 | EoG n=92 | EoD n=24 | EoG+EoD n=12 | |
|---|---|---|---|---|
|
| ||||
| Demographics & History: | ||||
|
| ||||
| Age at diagnosis: Median (IQR) | 10.6 (4.8–10.7) | 11.6 (6.7–14.7) | 7.4 (2.6–14.4) | 8.9 (5.2–13.2) |
| Male: n (%) | 74 (57.8%) | 48 (52.2%) | 18 (75.0%) | 8 (66.7%) |
| Atopic: n (%) | 91 (71.1%) | 62 (67.4%) | 19 (79.2%) | 10 (83.3%) |
| Asthma: n (%) | 44 (34.4%) | 30 (32.6%) | 11 (45.8%) | 3 (25.0%) |
| Allergic rhinitis: n (%) | 52 (40.6%) | 34 (37.0%) | 11 (45.8%) | 7 (58.3%) |
| Eczema: n (%) | 32 (25.0%) | 22 (23.9%) | 7 (29.2%) | 3 (25.0%) |
| Food allergy: n (%) | 49 (38.3%) | 35 (38.0%) | 10 (41.7%) | 4 (33.3%) |
| Previous diagnosis of EoE: n (%) | 23 (18.0%) | 16 (17.4%) | 5 (20.8%) | 2 (16.7%) |
|
| ||||
| Symptoms: | ||||
|
| ||||
| Number of symptoms per person: | 2.1 (1.0) | 2.0 (1.0) | 2.3 (1.1) | 1.8 (1.2) |
| Mean (SD) | ||||
| Dysphagia: n (%) | 25 (19.5%) | 18 (19.6%) | 6 (25.0%) | 1 (8.3%) |
| Feeding difficulty: n (%) | 21 (16.4%) | 13 (14.1%) | 5 (20.8%) | 3 (25.0%) |
| Abdominal pain: n (%) | 69 (53.9%) | 51 (55.4%) | 14 (58.3%) | 4 (33.3%) |
| Upper GI bleeding: n (%) | 6 (4.7%) | 3 (3.3%) | 2 (8.3%) | 1 (8.3%) |
| Nausea or vomiting: n (%) | 54 (42.2%) | 39 (42.4%) | 10 (41.7%) | 5 (41.7%) |
| Heartburn or regurgitation: n (%) | 16 (12.5%) | 15 (16.3%) | 1 (4.2%) | 0 |
| Diarrhea: n (%) | 21 (16.4%) | 14 (15.2%) | 5 (20.8%) | 2 (16.7%) |
|
| ||||
| Signs: | ||||
|
| ||||
| Anemia: n (%) | 15 (11.7%) | 12 (13.0%) | 2 (8.3%) | 1 (8.3%) |
| Hypoalbuminemia: n (%) | 2 (1.6%) | 1 (1.1%) | 1 (4.2%) | 0 |
| Peripheral eosinophils: Mean (SD) (cells/μL) |
700 (700) (n=67) |
500 (500) (n=46) |
800 (700) (n=12) |
900 (1400) (n=9) |
|
| ||||
| Gross findings: | ||||
|
| ||||
| Gastritis present: n (%) | 47 (36.7%) | 36 (39.1%) | 5 (20.8%) | 6 (50.0%) |
| Eosinophilic Gastritis Reference Score*: Mean (SD) |
1.2 (2.0) (n=109) |
1.2 (2.1) (n=79) |
0.3 (0.8) (n=19) |
1.5 (1.6) (n=11) |
| Duodenitis present: n (%) EoD Endo score: Mean (SD) |
34 (26.6%) 0.3 (0.6) |
10 (10.9%) 0.1 (0.4) |
15 (62.5%) 0.8 (0.8) |
9 (75.0%) 0.9 (0.7) |
| Visually normal: n (% ) | 37 (28.9%) | 32 (34.8%) | 4 (16.7%) | 1 (8.3%) |
|
| ||||
| Approach to Biopsy: | ||||
|
| ||||
| Patients with targeted biopsies*: n (%) | 53 (41.4%) | 27 (29.3%) | 16 (66.7%) | 10 (83.3%) |
| Number of gastric biopsies per person: Mean (SD) |
2.7 (1.1) | 2.8 (1.1) | 2.2 (1.1) | 2.7 (1.3) |
| Number of duodenal biopsies per person: Mean (SD) |
3.3 (1.7) | 2.9 (1.6) | 3.9 (1.7) | 4.4 (1.7) |
|
| ||||
| Histologic Findings: | ||||
|
| ||||
| Active EoE present: n (%) | 59 (46.1%) | 38 (41.3%) | 13 (54.2%) | 8 (66.7%) |
| Gastric eos/hpf: Mean (SD) | 88 (72) | 89 (73) | - | 74 (45) |
| Duodenal eos/hpf: Mean (SD) | 96 (53) | - | 108 (49) | 91 (42) |
Targeted biopsy: operative report noted a visual abnormality that was intentionally biopsied, the remainder of patients had random biopsies
Endoscopic and Histologic Features at Diagnosis
Gastric endoscopic abnormalities were present in 40% (n=42) of the 104 patients with EoG or EoG+EoD (Supplemental figure 1). The antrum was more affected than the body (EG-REFS scores 0.8 vs 0.4, p=0.006). The most common features were erythema, granularity, and erosions (Supplemental Table 1). In contrast, 66% (n=24) of patients with EoD or EoG+EoD had gross duodenitis. This predominantly affected the bulb (82%, n=19), and less commonly involved both the bulb and second portion (9%, n=2) or the second portion alone (4%, n=1) (p<0.0001). Gross duodenal abnormalities included erythema (50%, n=18), erosions (19%, n=7), salmon patches (8%, n=3), ulceration (6%, n=2), and villous atrophy (3%, n=1) (Supplemental figure 1). Over half (n=68) of EoG/EoD patients had visually normal mucosa in the setting of histologic inflammation.
In EoG and EoG+EoD, the most tissue eosinophils were in both the antrum and body in 48% (n=36), in the antrum alone in 35% (n=26), or the body alone in 17% (n=13) (p=0.003) (location not specified n=29). In EoD and EoG+EoD, the most eosinophils were in the bulb (63%, n=15), with fewer in the second portion (25%, n=6) or both the bulb and second portion (8%, n=2) (p=0.0004) (location not specified in n=12). Other histologic features were reported in the initial clinical pathology reports for 78% (n=81) of EoG patients, 89% (n=32) of EoD patients, and 56% (n=13) of low-grade eosinophilia patients (Supplemental Table 2, Supplemental Figure 2).
Initial Clinical Approaches
After diagnosis, 75% (n=96) of patients were treated for EoG or EoD. The most utilized were budesonide (65% granules, 35% crushed), PPI, systemic steroids, and dietary elimination therapy (DET)(Table 2). Combinations of the aforementioned treatments were utilized in 15% (n=19) and 9% (n=11) of patients were treated with esophageal topical corticosteroids (eTCS) for EoE in addition to their EoG/EoD treatment. There were no significant differences in initial treatment approach to EoG, EoD, or EoG+EoD (Table 2). However, not all patients were started on a new treatment after diagnosis: 14% continued previously prescribed PPIs, 9% were only treated with eTCS for EoE, and 2% were just observed (Table 2).
Table 2:
Follow-up, longitudinal trajectories, and disease complications
| EoG, EoD, or EoG+EoD: n=128 | EoG n=92 | EoD n=24 | EoG+EoD n=12 | |
|---|---|---|---|---|
|
| ||||
| Follow-up: | ||||
|
| ||||
| Follow-up in years: Mean (SD) | 5.8 (3.6) | 5.8 (3.8) | 5.6 (2.9) | 6.4 (3.3) |
| Patients with follow-up EGD‡: n (%) | 116 (90.6%) | 81 (88.0%) | 23 (95.8%) | 12 (100%) |
| Number of follow-up EGDs: Mean (SD) | 2.8 (2.6) | 2.8 (2.9) | 2.9 (1.6) | 3 (1.5) |
|
| ||||
| Initial treatment approach: n (%) | ||||
|
| ||||
| Treated for EoG or EoD: | 96 (75.0%) | 65 (70.6%) | 21 (87.5%) | 10 (83.3%) |
| Budesonide only | 41 (32.0%) | 26 (28.3%) | 10 (41.7%) | 5 (41.7%) |
| PPI only | 12 (9.4%) | 8 (8.7%) | 3 (12.5%) | 1 (8.3%) |
| Systemic steroids only | 11 (8.6%) | 8 (8.7%) | 1 (4.2%) | 2 (16.7%) |
| Dietary elimination therapy only | 13 (10.2%) | 7 (7.6%) | 5 (20.8%) | 1 (8.3%) |
| Combinations of the above | 19 (14.8%) | 16 (17.4%) | 2 (8.3%) | 1 (8.3%) |
| Not treated for EoG or EoD: | 32 (25.0%) | 27 (29.3%) | 3 (12.5%) | 2 (16.7%) |
| Continued pre-diagnosis PPI | 18 (14.1%) | 13 (14.1%) | 3 (12.5%) | 2 (16.7%) |
| Swallowed esophageal TCS | 11 (8.6%) | 11 (12.0%) | 0 | 0 |
| No treatment | 3 (2.3%) | 3 (3.3%) | 0 | 0 |
|
| ||||
| Longitudinal Trajectories: n (%) | ||||
|
| ||||
| EoG or EoD persisted/recurred: | 32 (25%) | 18 (19.6%) | 9 (37.5%) | 5 (41.7%) |
| Ultimately diagnosed with EoE: | 81 (63.3%) | 53 (57.6%) | 18 (75.0%) | 10 (83.3%) |
|
| ||||
| Complications: n (%) | ||||
|
| ||||
| All patients with complications | 23 (18.0%) | 14 (15.2%) | 6 (25.0%) | 3 (25.0%) |
| Moderate to severe anemia | 10 (7.8%) | 8 (8.7%) | 1 (4.2%) | 1 (8.3%) |
| Upper gastrointestinal bleeding | 9 (7.0%) | 5 (5.4%) | 3 (12.5%) | 1 (8.3%) |
| Hypoalbuminemia | 4 (3.1%) | 3 (3.3%) | 1 (4.2%) | 0 |
| Bowel perforation | 4 (3.1%) | 3 (3.3%) | 1 (4.2%) | 0 |
| Hospitalization | 4 (3.1%) | 3 (3.3%) | 0 | 1 (8.3%) |
| Surgery | 3 (2.3%) | 3 (3.3%) | 0 | 0 |
Natural History
Most patients (91%, n=116) underwent repeat endoscopy occurring on average 14.8 months after diagnosis (1.7 months-11.4 years) (Table 2). Only 25% (n=32) had persistent/recurrent eosinophilia (Figure 2a): half had persistent tissue eosinophilia on the first follow-up endoscopy (n=18) and half developed recurrent tissue eosinophilia later. The location of inflammation was relatively stable over time among EoG and EoD patients, whereas patients with EoG+EoD typically settled into either EoG or EoD (Figure 2b). Initial treatments were no different between patients with and without persistence/recurrence, and 88% patients were on a treatment for EoG or EoD immediately prior to the endoscopy demonstrating persistent/recurrent disease. Increases in EG-REFS scores were significantly associated with increased likelihood of a persistent/recurrent eosinophilia (aOR 1.34, CI 1.03–1.74) (Table 3). No differences were identified in baseline EoG histologic scores when comparing those with and without persistent/recurrent eosinophilia (0.3 vs 0.2, p=0.06) (Supplemental Table 3).
Figure 2:
Natural History in EoG and EoD
A. Proportion of EoG/EoD patients with persistence/recurrence of eosinophilia vs proportion of low-grade eosinophilia patients who later developed EoG/EoD
B. Heat map illustrating patterns of histologic disease activity over repeat endoscopies
Table 3:
Associations between features at diagnosis and persistent/recurrent eosinophilia or late complications
| Features at diagnosis | Persistent/ recurrent eosinophilia n=32 | No persistent/ recurrent eosinophilia n=96 | Univariate logistic regression OR (95% CI) | Multivariate logistic regression** aOR (95% CI) AUC=0.75 | Late complication n=11 | No late complication n=112 | Univariate logistic regression OR (95% CI) | Multivariate logistic regression** aOR (95% CI) AUC =0.95 |
|---|---|---|---|---|---|---|---|---|
| Age at diagnosis (years)Δ: Mean (SD) | 10.3 (4.8) | 9.6 (5.8) | 1.00 (1.00–1.01) | 8.5 (3.6) | 9.9 (5.8) | 1.00 (0.99–1.01) | ||
| Male sex#: n (%) | 20 (66.7%) | 54 (56.3%) | 1.30 (0.58–3.01) | 8 (72.7%) | 66 (58.9%) | 2.06 (0.52–8.16) | ||
| Complication present at diagnosis#: n (%) | 7 (21.9%) | 8 (8.3%) | 3.08 (1.02–9.32)* | 3.65 (0.57–23.52) | 4 (36.4%) | 11 (9.8%) | 5.51 (1.39–21.81)* | 9.63 (1.09–85.20)* |
| Peripheral Eosinophil Count†: Mean (SD) | 0.77 (1.01) | 0.62 (0.59) | 1.01 (0.99–1.03) | 1.12 (1.37) | 0.59 (0.56) | 1.01 (0.99–1.03) | ||
| EG-REFS score‡: Mean (SD) | 2.00 (2.81) | 0.94 (1.70) | 1.28 (1.05–1.57)* | 1.34 (1.03–1.74)* | 3.25 (3.32) | 1.02 (1.84) | 1.52 (1.14–2.03)* | 1.70 (1.11–2.63)* |
| EoD Endo score‡: Mean (SD) | 0.5 (0.7) | 0.3 (0.6) | 1.85 (1.01–3.38)* | 1.72 (0.56–5.32) | 0.7 (1.0) | 0.3 (0.6) | 2.28 (1.04–4.98)* | 8.74 (1.67–45.60)* |
| Gastric eosinophilia‖ present#: n (%) | 23 (71.9%) | 81 (84.4%) | 0.47 (0.19–1.25) | 8 (72.7%) | 96 (85.7%) | 0.58 (0.15–2.83) | ||
| Duodenal eosinophilia¶ present#: n (%) | 14 (43.8%) | 22 (22.9%) | 2.62 (1.12–6.10)* | 5.9 (0.60–23.52) | 5 (45.4%) | 31 (27.7%) | 2.31 (0.66–8.12) | |
| Eos/hpf in stomach: Mean (SD) | 116 (87) | 81 (64) | 1.12 (1.00–1.28)* | 1.0 (0.89–1.26) | 135 (82) | 85 (69) | 1.15 (0.99–1.34) | |
| Eos/hpf in duodenum: Mean (SD) | 137 (48) | 78 (29) | 2.25 (1.33–3.80)* ◊ | 131 (73) | 96 (41) | 1.32 (0.90–1.92) | ||
| Follow up time (years)Δ: Mean (SD) | 7.4 (4.1) | 5.4 (3.3) | 1.12 (1.01–1.25)* | 1.11 (0.93–1.32) | 7.3 (4.0) | 5.6 (3.4) | 1.22 (1.05–1.41)* | 1.17 (0.90–1.50) |
Odds ratios reflect changes per 1 year increase.
Odds ratios reflect the presence of this feature.
Odds ratios reflect changes per 20 unit increase in eosinophils.
Odds ratios reflect changes per 1 unit increase in score.
≥30 eos/hpf in stomach.
≥50 eos/hpf in duodenum.
p<0.05.
features significant on univariate analysis were included in multivariate analysis.
Feature not included in presented multivariate model due to collinearity with “Duodenal eosinophilia present” and high degree of missing data. In a separate model including eos/hpf in duodenum in place of “Duodenal eosinophilia present” it still was not significant on multivariate modeling.
The majority (75%, n=96) of patients had no recurrence of mucosal eosinophilia. While 15% (n=14) of these had normal follow-up endoscopies off of any EGID treatment, 73% (n=70) were still being treated for EGID at the time of the last endoscopic evaluation, and 11% (n=11) were managed without follow-up endoscopy.
Disease Complications
Disease complications occurred in 18% (n=23) of patients (Table 2, Figure 3a). Among patients with complications, most (70%, n=16) had a single complication. Patients with multiple complications (30%, n=7) experienced 2–10 complications per person (Figure 3c). Complications occurred at diagnosis in 30% and later in the course in 70%, at an average of 3.7 years after diagnosis (0.3–11.5 years) (Figure 3b). The most common were anemia (8%, n=10) and UGIB (7%, n=9). The average hemoglobin among patients with anemia without UGIB was 9.5g/dL (SD 1.0 mg/dL), and patients with UGIB had an average of 7.9mg/dL (SD 1.7mg/dL). Treatments for UGIB included hospitalization and supportive care (92%), red blood cell transfusion (61%), endoscopy (70%), and endoscopic control of bleeding (23%). Albumin levels among patients with hypoalbuminemia were 2.74g/dL (SD 0.38g/dL). Surgeries included repair of bowel perforation (n=4), partial gastrectomy for nonhealing ulcers (n=3), and pyloroplasty for pyloric stenosis (n=1). No significant differences were found in frequency or type of complications between patients with EoG, EoD, or EoG+EoD (Table 2), or the type of complications occurring at or after diagnosis. However, the presence of a complication at diagnosis (aOR 9.63, CI 1.09–85.20), increases in EoD-Endo score (aOR 8.74, CI 1.67–45.60), and increases in EG-REFS scores (aOR 1.70, CI 1.11–2.63) were associated with increased likelihood of late complications (Table 3).
Figure 3:
Disease Complications in EoG and EoD
A. Proportion of EoG/EoD patients with complications vs patients with low-grade eosinophilia
B. Types of complications and timing relative to diagnosis
C. Heat map illustrating type and timing of late complications
Special Subpopulations
Patients with low-grade gastric or duodenal eosinophilia
Twenty-three patients with low-grade gastric and/or duodenal eosinophilia were identified. Most of these patients were treated (60%, n=14), and the rest were observed (21%, n=5) or were treated only for concomitant EoE (17%, n=4). Follow-up endoscopy was performed in 91% (n=21) of these patients, with an average of 3.7 (SD 2.9) follow-up endoscopies per person. Only one patient went on to develop EoG (90eos/hpf), which was noted on the fourth endoscopic evaluation two years after initial presentation (Figure 2a). Fifty six percent (n=13) had a diagnosis of EoE by the end of the follow-up period. None developed disease complications (Figure 3a).
Patients with concomitant EoE
Fifty one percent (n=65) of patients were diagnosed with EoE before or at the time of EoG/EoD diagnosis. These patients were more likely to be male (75% vs 40%, p <0.0001), have a higher burden of atopic comorbidities (2.5±1.5 vs 1.3±1.3, p<0.0001), and have higher peripheral eosinophil counts (834±826 vs 426±499 cells/μL, p=0.02). No differences were found in likelihood of developing complications (15% vs 19%, p=0.6) or recurrent/persistent eosinophilia (25% vs 25%, p=0.9) when comparing those with EoE and those without. By the end of the follow-up period, 63% (n=80) of the cohort was diagnosed with EoE.
Discussion:
In this largest and longest natural history study of children with gastric or duodenal eosinophilia we report a number of novel and clinically impactful findings. First, we found that the annual diagnoses of EoG and EoD increased during the study. Second, with an average follow-up time of 5.8 years, gastroduodenal eosinophilia persisted or recurred in 25% of patients. Third, 18% of children suffered disease complications including recurrent bleeding, perforation, and anemia. Finally, the severity of endoscopic abnormalities at diagnosis was found to associate with long term outcomes, whereas peak eosinophil counts did not. Together, these findings provide guidance for disease detection and long-term follow-up.
After systematically identifying pediatric patients with EoG or EoD we noted an upward trend in the number of patients diagnosed annually, similar to previous work.9 This persisted after adjusting for the number of patients undergoing endoscopy annually. In EoE, increasing disease incidence also outpaces increases in endoscopy volume, and a combination of determinants such as early life exposures, genetic predisposition, and environmental influences are suspected.25,26 The similar, albeit less prodigious, increase in diagnosis of pediatric EoG and EoD reported here could reflect similar causal elements, and merits further study. Unexpectedly, we found high rates of misclassification using SNOMED CT codes for EoG and EoD. This should be considered when interpreting the results of administrative claims studies of non-esophageal EGIDs that do not incorporate histologic review, and validating administrative claims codes for EoG and EoD should be prioritized.
Our finding that one quarter of children with EoG or EoD had persistent/recurrent eosinophilia over time builds on previous work exploring symptomatic trajectories in children and adults. This is lower than the 58–75% of patients with relapsing or continuous symptoms in Pineton de Chambrun et al’s landmark natural history study and others describing longitudinal symptom patterns. 11,22,27 However, Reed et al reported that while histologic and endoscopic remission are highly concordant in patients with EoG/EoD, 37% experience persistent symptoms despite histologic remission defined by peak eosinophil count.28 Biologic therapies that successfully reduce or deplete tissue eosinophils such as reslizumab, mepolizumab, or benralizumab have not consistently demonstrated symptomatic improvements among EGID patients. 29–31 Future prospective studies should explore the longitudinal associations between symptoms and markers of disease activity, with markers of disease activity including endoscopic and histologic measures in addition to peak eosinophil counts.
We identified a complication rate of 18% among pediatric patients with EoG or EoD. We note specific complication frequencies similar to previous studies in children and adults with regards to anemia (5–9%)13,27,28,32, UGIB (6–9%)4,33, and hypoalbuminemia (3%)27, though some studies do report higher frequencies.3,12,13,28,32 However, the literature to date on this topic is heterogeneous and utilizes variable nomenclatures (“eosinophilic gastroenteritis,” not indicating segments involved), disease definitions (no eos/hpf thresholds or more permissive thresholds: 20eos/hpf), and study designs (case series). Utilizing the new consensus guidelines for EGID diagnosis will facilitate direct comparison of different studies’ results and will help us better characterize any differences that may exist between EoG and EoD.15,16 Of note, in our cohort the thresholds set forth in these guidelines had a high sensitivity for identifying patients who developed either downstream complications or recurrent tissue eosinophilia.
Additionally, we identified a number of novel features associated with persistent/recurrent tissue eosinophilia or downstream complications including presenting with a complication and the extent of endoscopic abnormalities at diagnosis in either the stomach or duodenum. Specifically, our results support the clinical usefulness of EG-REFS scoring, noting an increase in the likelihood of persistent/recurrent eosinophilia or downstream complications with higher composite scores. This is meaningful because in both our cohort and that of Pesek et al, 62–63% of pediatric patients with EoG had visually normal gastric tissue at diagnosis, and Hirano et al found that gastric peak eosinophil counts only moderately correlated with EG-REFS scores (r=0.32–0.43).17,23 Our data suggest that practical measures like assessment of endoscopic severity carry more weight than eosinophil enumeration in parsing out the patients at risk for a more severe disease course. Patients presenting with gastroduodenal endoscopic abnormalities or complications should be followed closely.
While our study provides important novel findings, limitations are associated with its retrospective design. We were unable to reliably assess symptom changes over time and do not account for patients with histologic quiescence but ongoing symptoms. Additionally, in this real-world cohort only 15% of patients were evaluated for remission off of treatment. Other patients might have had recrudescent eosinophilia on a treatment-free reevaluation. Despite these limitations, our findings are relevant and important for the identification and care of EGID patients, especially those with more severe disease. For instance, 88% of the patients with persistent/recurrent mucosal eosinophilia were also on treatment at the time of the endoscopy demonstrating persistence/recurrence. With the advent of biologics targeting Th2 and eosinophil-related pathways, earlier or more aggressive treatments may be valuable in some patients.
Since our work focused on children from North America, it provides an important contrast to the robust literature in Japan where EoG and eosinophilic enteritis are more common than EoE.11,34,35 Differences in dairy consumption or genetic differences could influence these variable phenotypes. Previous work by Shoda et al noted slight differences in EoE biopsy transcriptomes in Japanese patients, such as lower thymic stromal lymphopoietin levels and higher filaggrin levels, but population-based differences in EoG/EoD genetic profiles have not been explored.36 Delving further into epidemiologic or genetic determinants in these populations may provide a window into the pathomechanisms of EGIDs.
In conclusion, we demonstrate that complications or persistence/recurrence of gastroduodenal eosinophilia affect up to 1/4 of pediatric patients with EoG or EoD. Patients with endoscopic abnormalities or complicated clinical presentations may be at risk for more severe disease and should be followed closely. Future studies should seek to identify clinical or molecular features that prospectively associate with long term outcomes in these heterogeneous disorders.
Supplementary Material
Study Highlights:
WHAT IS KNOWN:
Patients with eosinophilic gastritis or duodenitis have unique longitudinal trajectories and disease complications that differ from eosinophilic esophagitis.
The frequency of histologic recurrence or disease complications in eosinophilic gastritis and duodenitis is not well-described.
WHAT IS NEW HERE:
Gastroduodenal eosinophilia is increasingly being recognized in children.
A minority of children with gastroduodenal eosinophilia experience recurrence or disease complications, but certain baseline features associate with long term outcomes.
Study Support:
T32DK067009, Colorado CTSA Grant Number UL1TR002535 – LAQ, K23DK109263-05 – CMK, LaCache Chair in Gastrointestinal Immunologic and Allergic Diseases – GTF.
Footnotes
Conflicts of Interest
The authors have no conflicts of interest to disclose regarding this work with the exception of the research funding outlined above.
References:
- 1.Kaijser R. Zur Kenntnis der allergischen Affektionen des Verdauungakan als vom Standpunkt des Chirurgen aus. Arch Klin Chir. 1937;188: 36–64. [Google Scholar]
- 2.Klein NC, Hargrove RL, Sleisenger MH, et al. Eosinophilic gastroenteritis. Medicine. 1970;49: 299–319. [DOI] [PubMed] [Google Scholar]
- 3.Mansoor E, Saleh MA, Cooper GS. Prevalence of Eosinophilic Gastroenteritis and Colitis in a Population-Based Study, From 2012 to 2017. Clin Gastroenterol Hepatol. 2017;15: 1733–1741. [DOI] [PubMed] [Google Scholar]
- 4.Jensen ET, Martin CF, Kappelman MD, et al. Prevalence of Eosinophilic Gastritis, Gastroenteritis, and Colitis: Estimates From a National Administrative Database. J Pediatr Gastroenterol Nutr. 2016;62: 36–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Shoda T, Wen T, Caldwell JM, et al. Molecular, endoscopic, histologic, and circulating biomarker-based diagnosis of eosinophilic gastritis: Multi-site study. J Allergy Clin Immunol. 2020;145: 255–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Caldwell JM, Collins MH, Stucke EM, et al. Histologic eosinophilic gastritis is a systemic disorder associated with blood and extragastric eosinophilia, TH2 immunity, and a unique gastric transcriptome. J Allergy Clin Immunol. 2014;134: 1114–1124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ben-Baruch Morgenstern N, Shoda T, Rochman Y, et al. Local type 2 immunity in eosinophilic gastritis. J Allergy Clin Immunol. 2023. doi: 10.1016/j.jaci.2023.01.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Shoda T, Rochman M, Collins MH, et al. Molecular analysis of duodenal eosinophilia. J Allergy Clin Immunol. 2022. doi: 10.1016/j.jaci.2022.12.814 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Pesek RD, Reed CC, Muir AB, et al. Increasing Rates of Diagnosis, Substantial Co-Occurrence, and Variable Treatment Patterns of Eosinophilic Gastritis, Gastroenteritis, and Colitis Based on 10-Year Data Across a Multicenter Consortium. Am J Gastroenterol. 2019;114: 984–994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Pineton de Chambrun G, Dufour G, Tassy B, et al. Diagnosis, Natural History and Treatment of Eosinophilic Enteritis: a Review. Curr Gastroenterol Rep. 2018;20: 37. [DOI] [PubMed] [Google Scholar]
- 11.Yamamoto M, Nagashima S, Yamada Y, et al. Comparison of Nonesophageal Eosinophilic Gastrointestinal Disorders with Eosinophilic Esophagitis: A Nationwide Survey. J Allergy Clin Immunol Pract. 2021. doi: 10.1016/j.jaip.2021.06.026 [DOI] [PubMed] [Google Scholar]
- 12.Grandinetti T, Biedermann L, Bussmann C, et al. Eosinophilic Gastroenteritis: Clinical Manifestation, Natural Course, and Evaluation of Treatment with Corticosteroids and Vedolizumab. Dig Dis Sci. 2019;64: 2231–2241. [DOI] [PubMed] [Google Scholar]
- 13.Chehade M, Magid MS, Mofidi S, et al. Allergic eosinophilic gastroenteritis with protein-losing enteropathy: intestinal pathology, clinical course, and long-term follow-up. J Pediatr Gastroenterol Nutr. 2006;42: 516–521. [DOI] [PubMed] [Google Scholar]
- 14.Egan M, Furuta GT. Eosinophilic gastrointestinal diseases beyond eosinophilic esophagitis. Ann Allergy Asthma Immunol. 2018;121: 162–167. [DOI] [PubMed] [Google Scholar]
- 15.Dellon ES, Gonsalves N, Abonia JP, et al. International Consensus Recommendations for Eosinophilic Gastrointestinal Disease Nomenclature. Clin Gastroenterol Hepatol. 2022;20: 2474–2484.e3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Papadopoulou A, Amil-Dias J, Auth MK-H, et al. Joint ESPGHAN/NASPGHAN Guidelines on Childhood Eosinophilic Gastrointestinal Disorders beyond Eosinophilic Esophagitis. J Pediatr Gastroenterol Nutr. 2023. doi: 10.1097/MPG.0000000000003877 [DOI] [PubMed] [Google Scholar]
- 17.Pesek RD, Reed CC, Collins MH, et al. Association Between Endoscopic and Histologic Findings in a Multicenter Retrospective Cohort of Patients with Non-esophageal Eosinophilic Gastrointestinal Disorders. Dig Dis Sci. 2020;65: 2024–2035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Di Ruscio M, Variola A, Vernia F, et al. Role of Ulcerative Colitis Endoscopic Index of Severity (UCEIS) versus Mayo Endoscopic Subscore (MES) in Predicting Patients’ Response to Biological Therapy and the Need for Colectomy. Digestion. 2021;102: 534–545. [DOI] [PubMed] [Google Scholar]
- 19.Xie T, Zhang T, Ding C, et al. Ulcerative Colitis Endoscopic Index of Severity (UCEIS) versus Mayo Endoscopic Score (MES) in guiding the need for colectomy in patients with acute severe colitis. Gastroenterol Rep. 2018;6: 38–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Allez M, Lemann M, Bonnet J, et al. Long term outcome of patients with active Crohn’s disease exhibiting extensive and deep ulcerations at colonoscopy. Am J Gastroenterol. 2002;97: 947–953. [DOI] [PubMed] [Google Scholar]
- 21.Laserna-Mendieta EJ, Casabona S, Guagnozzi D, et al. Efficacy of proton pump inhibitor therapy for eosinophilic oesophagitis in 630 patients: results from the EoE connect registry. Aliment Pharmacol Ther. 2020;52: 798–807. [DOI] [PubMed] [Google Scholar]
- 22.Wolf WA, Cotton CC, Green DJ, et al. Predictors of response to steroid therapy for eosinophilic esophagitis and treatment of steroid-refractory patients. Clin Gastroenterol Hepatol. 2015;13: 452–458. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hirano I, Collins MH, King E, et al. Prospective Endoscopic Activity Assessment for Eosinophilic Gastritis in a Multisite Cohort. Am J Gastroenterol. 2022;117: 413–423. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.World Health Organization. Iron deficiency anemia. Assessment, prevention, and control. A guide for programme managers. 2001; 47–62. [Google Scholar]
- 25.Dellon ES, Erichsen R, Baron JA, et al. The increasing incidence and prevalence of eosinophilic oesophagitis outpaces changes in endoscopic and biopsy practice: national population-based estimates from Denmark. Aliment Pharmacol Ther. 2015;41: 662–670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Navarro P, Arias Á, Arias-González L, et al. Systematic review with meta-analysis: the growing incidence and prevalence of eosinophilic oesophagitis in children and adults in population-based studies. Aliment Pharmacol Ther. 2019;49: 1116–1125. [DOI] [PubMed] [Google Scholar]
- 27.Pineton de Chambrun G, Gonzalez F, Canva J-Y, et al. Natural history of eosinophilic gastroenteritis. Clin Gastroenterol Hepatol. 2011;9: 950–956.e1. [DOI] [PubMed] [Google Scholar]
- 28.Reed C, Woosley JT, Dellon ES. Clinical characteristics, treatment outcomes, and resource utilization in children and adults with eosinophilic gastroenteritis. Dig Liver Dis. 2015;47: 197–201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kliewer KL, Murray-Petzold C, Collins MH, et al. Benralizumab for eosinophilic gastritis: a single-site, randomised, double-blind, placebo-controlled, phase 2 trial. Lancet Gastroenterol Hepatol. 2023;8: 803–815. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Dellon ES, Peterson KA, Mitlyng BL, et al. Mepolizumab for treatment of adolescents and adults with eosinophilic oesophagitis: a multicentre, randomised, double-blind, placebo-controlled clinical trial. Gut. 2023;72: 1828–1837. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Spergel JM, Rothenberg ME, Collins MH, et al. Reslizumab in children and adolescents with eosinophilic esophagitis: results of a double-blind, randomized, placebo-controlled trial. J Allergy Clin Immunol. 2012;129: 456–63, 463.e1–3. [DOI] [PubMed] [Google Scholar]
- 32.Ko HM, Morotti RA, Yershov O, et al. Eosinophilic gastritis in children: clinicopathological correlation, disease course, and response to therapy. Am J Gastroenterol. 2014;109: 1277–1285. [DOI] [PubMed] [Google Scholar]
- 33.Chehade M, Kamboj AP, Atkins D, et al. Diagnostic Delay in Patients with Eosinophilic Gastritis and/or Duodenitis: A Population-Based Study. J Allergy Clin Immunol Pract. 2021. doi: 10.1016/j.jaip.2020.12.054 [DOI] [PubMed] [Google Scholar]
- 34.Kobayashi S, Tsunoda T, Umetsu S, et al. Clinical features of pediatric eosinophilic gastroenteritis. Pediatr Int. 2022;64: e15322. [DOI] [PubMed] [Google Scholar]
- 35.Nomura I, Morita H, Ohya Y, et al. Non-IgE-mediated gastrointestinal food allergies: distinct differences in clinical phenotype between Western countries and Japan. Curr Allergy Asthma Rep. 2012;12: 297–303. [DOI] [PubMed] [Google Scholar]
- 36.Shoda T, Morita H, Nomura I, et al. Comparison of gene expression profiles in eosinophilic esophagitis (EoE) between Japan and Western countries. Allergol Int. 2015;64: 260–265. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.