Skin barrier function in children with eosinophilic esophagitis

Abstract

Background

Eosinophilic esophagitis (EoE) is associated with epithelial barrier dysfunction of the esophagus. Studies suggest that the skin lipidomic profile is altered in EoE; therefore, we hypothesized that skin barrier function would be compromised in subjects with EoE.

Objective

We aimed to evaluate skin barrier function in subjects with EoE by assessing transepidermal water loss (TEWL).

Methods

We performed a single-center, case-control study of pediatric subjects with known EoE and controls without EoE. The subjects with EoE met consensus criteria for diagnosis. Patients receiving systemic biologic medications or prescription topical corticosteroids were excluded. A majority of the subjects with EoE underwent a concurrent esophageal string test or had a recent endoscopy near the time of TEWL assessment. In addition to comparing subjects with EoE and controls without EoE, we compared TEWL according to EoE disease activity.

Results

The subjects with EoE (n = 25) were older than the controls without EoE (n = 24) but similar from the standpoint of other demographic features. The controls were primarily atopic (79.2%). Rate of TEWL was comparable in the subjects with EoE and the controls (median =15.1 [interquartile range = 4.95] g/m² per hour vs median = 16.9 [interquartile range = 7.3] g/m² per hour [P = .18]); differences between those subjects with active EoE (n = 10) and those with inactive EoE (n = 14) were not observed. A subgroup analysis excluding patients with atopic dermatitis did not reveal differences in TEWL among the subjects with EoE.

Conclusions

Subjects with EoE did not have marked skin barrier dysfunction relative to the controls without EoE and TEWL did not vary according to disease activity. These findings suggest that the utility of TEWL as a biomarker of EoE is limited. Larger studies with defined atopic and nonatopic controls are needed to detect small to moderate differences in skin barrier function.

Introduction

Eosinophilic esophagitis (EoE) is a chronic allergic disease associated with tissue eosinophilia and symptoms of esophageal dysfunction. The presence of dilated intercellular spaces and downregulation of barrier proteins (eg, filaggrin, zonulin-1) and adhesion molecules (eg, desmoglein-1) suggest that EoE is associated with impaired esophageal epithelial integrity.¹ Further, lipidomic analysis has recently revealed distinct abnormalities in the sphingolipid composition of brushings from the esophagus.² Intriguingly, Goleva et al³ recently reported alterations in the lipidomic profile of the skin of pediatric subjects with EoE, suggesting that epithelial barrier dysfunction in EoE may not be limited to the esophagus. The authors demonstrated reduced production of long-chain base ceramides, especially ultra-long-chain fatty acids. Although these contribute to the barrier properties of the skin, the functional relevance of these findings has not been examined.

Transepidermal water loss (TEWL) quantifies passive diffusion of water vapor from the viable epidermis through the stratum corneum and has been used to assess skin barrier function in allergic disease. Relative to healthy children, TEWL is increased in subjects with asthma, allergic rhinitis, atopic dermatitis (AD), and food allergy.⁴ Moreover, increased TEWL in early life may predict the onset of allergic sensitization to foods.⁵ Herein, we have used TEWL to assess the skin barrier function of subjects with EoE.

Results and discussion

We performed a pilot study comparing TEWL in pediatric subjects with EoE (n = 25) and controls without EoE (n = 24) (Phoenix Children’s institutional review board approval no. 23-213). Subjects were recruited from the Allergy/Immunology and Gastroenterology Clinics at Phoenix Children’s. The participants with EoE met consensus diagnostic criteria.⁶ Patients without a history of EoE were recruited as controls. Most subjects with EoE also underwent a concurrent esophageal string test to assess disease activity (n = 20)⁷; the remaining subjects (n = 4 or 5) had a recent endoscopy. Patients requiring systemic (ie, biologics and Janus kinase inhibitors) or topical (ie, glucocorticoids, phosphodiesterase 4 inhibitors, and calcineurin inhibitors) therapy within 6 months or reporting any emollient use within 1 week were excluded. Caffeine consumption within 1 day was also an exclusion criterion owing to potential interference with TEWL. TEWL was measured twice on the volar forearm by using the Aquaflux Model AF200 (BioX Systems Ltd, London, United Kingdom), and the measurements were averaged. Atopic disease status was based on physician diagnosis of asthma, eczema, allergic rhinitis, or IgE-mediated food allergy by using clinical symptoms and testing as indicated (eg, skin testing, serum IgE level, spirometry, or supervised challenge) and was extracted from the electronic medical record. Demographic characteristics and atopic disease frequencies were compared by using a chi-square test or the Fisher exact test. The Mann-Whitney U test was used to compare age and TEWL between groups.

The characteristics of the study participants are detailed in Table I. The subjects with EoE were older than the controls (P = .006) but similar from the standpoint of other demographics, including sex, race, and ethnicity. Subjects with EoE were more likely to have IgE-mediated food allergy and AD and less likely to have asthma or allergic rhinitis, although the differences between groups were not significant for these features. Of the controls without EoE, 79.2% (n = 19) were atopic and 20.8% (n = 4) were nonatopic. TEWL was comparable in the controls and the subjects with EoE (median =15.1 [interquartile range (IQR) = 4.95] vs 16.9 [IQR = 7.3] g/m² per hour] [P = .18]) (Fig 1, A) and in subjects with inactive (n = 14) vs active EoE (n = 10) (median =18.5 [IQR = 8.7] vs 16.9 [IQR = 4.2] g/m² per hour [P = .63]) (Fig 1, B). An exploratory subgroup analysis excluding all patients with AD did not reveal differences in TEWL among the subjects with EoE (n = 18) versus TEWL among the controls (n = 21) (median = 17.8 [IQR = 7.3] vs 14.8 [IQR = 4.6] g/m² per hour [P = .08]) (Fig 1, C).

Table I.

Characteristics of the study population

Characteristic	Atopic controls (n = 24)	Subjects with EoE (n = 25)
Age (y), median (IQR)∗	11.0 (6.5)	13.8 (4.1)
Sex, no. (%)
Female	12 (50)	7 (28)
Male	12 (50)	18 (72)
Race and ethnicity, no. (%)
Hispanic	5 (20.8)	4 (16)
White	14 (58.3)	19 (76)
Other	5 (20.8)	2 (8)
Atopic conditions, no. (%)
Allergic rhinitis	17 (70.8)	15 (60)
Asthma	10 (41.7)	5 (20)
AD	3 (12.5)	7 (28)
Food allergy	3 (12.5)	10 (40)

^∗P =.006 (according to the Mann-Whitney U test).

Fig 1.

TEWL in the controls without EoE versus in the subjects with EoE (A), subjects with inactive versus active EoE (B), and controls without EoE and AD versus in subjects with EoE and without AD (C).

The median TEWL values observed in both the controls and subjects with EoE were higher than expected (6-10 g/m² per hour) for healthy children and adolescents aged 6 to 17 years.⁸ This is likely because our control group predominantly comprised patients with atopy.⁴ To our knowledge, only 1 other study has investigated TEWL in subjects with EoE.⁹ Differences in TEWL were not observed when subjects with EoE were compared with controls without EoE; however, whether patients with AD had been excluded is not clear. The cohort was also smaller (n = 17 subjects with EoE vs 9 controls without EoE) and included pediatric and adult subjects.

In terms of limitations, our study was a small, single-center study of pediatric subjects; therefore, the results cannot be generalized to adults. Disease activity in subjects with EoE was defined primarily by the esophageal sting test rather than by endoscopic biopsy. TEWL was not examined exclusively in nonatopic controls, although the prevalence of comorbid atopy in EoE is 60% to 80%,¹⁰ suggesting that our control group was representative of patients who may be evaluated for EoE in an allergy clinic. Our sample was not powered to detect small to moderate differences in TEWL, and unlike Goleva et al,³ we did not age match or exclude all patients with AD. In our study, the subjects with EoE were older, and we observed a trend toward increased TEWL in EoE when all subjects with AD were strictly excluded. TEWL decreases throughout childhood¹¹; therefore, age matching could magnify potential differences between subjects with EoE and atopic controls.

Our results suggest that skin barrier dysfunction is not markedly increased in patients with EoE relative to the patients commonly seen in allergy clinics. On the basis of expected published values, TEWL may be increased in patients with EoE, which is similar to TEWL observed in patients with other atopic diseases. Epicutaneous antigen sensitization to EoE food triggers may occur through the skin. Animal models suggest that immune cell trafficking and systemic inflammatory mediator release (eg, IL-33, TSLP) along the skin-gut axis promotes gastrointestinal eosinophilia.¹² Why some patients with atopic disease and elevated TEWL may be more prone to cutaneous sensitization is still unknown. Our findings suggest that TEWL has limited utility in discriminating patients with EoE. Moreover, we did not observe differences in TEWL according to disease activity, suggesting that skin barrier dysfunction is unaffected by EoE treatments. Of note, patients treated with biologics or Janus kinase inhibitors were excluded. Integration of TEWL and electric impedance spectroscopy¹³ may be useful to evaluate skin barrier function in EoE in the future.

Key messages.

• •TEWL has limited utility as a biomarker in EoE.
• •Additional studies are needed to evaluate skin barrier function in EoE and potential mechanisms of food antigen sensitization along the skin-gut axis in EoE.

Disclosure statement

Supported by the Phoenix Children’s Foundation and, in part, by the National Institutes of Health/National Institute of Allergy and Infectious Diseases (grant K23AI158813 [to B.L.W.])

Disclosure of potential conflict of interest: S. Schroeder reports serving as a consultant for Regeneron, Sanofi, Takeda, and EvoEndo and receiving research funding from Regeneron, Celgene, and AstraZeneca. C. S. Bauer reports receiving research funding from OM Pharma, AstraZeneca, and Sanofi. B. L. Wright reports receiving in-kind support from Regeneron in the form of study drug (dupilumab and placebo) for a clinical trial of milk oral immunotherapy. The rest of the authors declare that they have no relevant conflicts of interest.