INTRODUCTION

Treatments for eosinophilic esophagitis (EoE), a chronic allergic disease that is increasing in incidence and prevalence, consist of both pharmacologic and nonpharmacologic approaches (1,2). Primary pharmacologic approaches include proton-pump inhibitors, topical corticosteroids (tCS), and biologics (3–6). Because there are no comparative effectiveness trials of these medication classes, selection of initial therapy is generally empiric and based on a shared decision-making model taking into account patient and provider preferences and access to specific treatments (7). In practice, tCS are among the most commonly used medications and have received a “strong” recommendation in recent guidelines (3,8,9). However, nonresponse to tCS has been reported to be as high as 40%, and despite active investigation into predictors of response to tCS, no clinical or molecular marker has yet to be validated (10–17). Nevertheless, identifying biomarkers that predict treatment response or nonresponse to tCS for patients with EoE would accelerate patient care because nonresponders could be directed to alternative treatment strategies, with an ultimate goal of improving outcomes and decreasing the chance for disease progression (18,19).

DNA methylation can occur at specific DNA sequences (cytosines followed by guanine bases, or CpG sites) and is important in gene regulation. Often, increased DNA methylation is associated with reduced gene expression, but the direction of the regulation is site specific. The degree of methylation can be influenced by the environmental factors, including environmental factors directly or indirectly related to EoE treatment response. We hypothesized that pretreatment DNA methylation patterns in esophageal tissue could partially explain heterogeneity in response to topical steroid treatment, and thus, a pretreatment epigenetic signature could serve as a novel biomarker for response to treatment. Using archived baseline (pretreatment) esophageal biopsies, we previously identified differential methylation at 18 CpG sites in pretreatment samples when comparing treatment responders with nonresponders (20). While promising, these findings needed to be validated in an independent sample before clinical use. The aim of this study, therefore, was to test for association between DNA methylation at these 18 previously identified CpG sites that exhibited differential DNA methylation between responders and nonresponders and to validate their utility as biomarkers for treatment response to topical steroids for patients diagnosed with EoE.

METHODS Parent study design, patients, samples, and outcomes

We analyzed data and biospecimens from a randomized, double-blind, double-dummy, clinical trial of budesonide vs fluticasone (NCT02019758) (17,21,22). The trial enrolled patients with a new diagnosis of EoE who were then randomized to receive either active oral viscous budesonide (1 mg twice a day) plus a placebo inhaler or fluticasone from a multidose inhaler (880 mcg twice a day) or a placebo slurry. Patient outcomes were assessed after an 8-week treatment course. Because the trial demonstrated no significant differences in response between the topical steroid formulations (22), this study included either tCS treatments for the prediction of response.

Upper endoscopy was performed at baseline (pretreatment) and after 8 weeks of treatment. During both endoscopies, endoscopic severity was assessed with the EoE Endoscopic Reference Score (EREFS) (23–25), and biopsies were obtained to determine the peak esophageal eosinophil count (eosinophils per high-power field [eos/hpf]), quantified as per our previously validated method with a microscope hpf size of 0.24 mm (26,27). An additional biopsy was obtained at baseline from the mid-esophagus (approximately 10 cm above the gastroesophageal junction), flash frozen in liquid nitrogen, and then stored at −80 oC. We used the mid-esophageal location for this analysis based on prior results showing uniform gene expression throughout the esophagus in EoE (28,29).

We defined the outcome of histologic response as a peak posttreatment count of <15 eos/hpf across all biopsy sites at repeat endoscopy after 8 weeks of treatment; nonresponders were defined as a peak posttreatment count of ≥15 eos/hpf (30–32). This study was approved by the University of North Carolina and Wake Forest Institutional Review Boards, and patients previously provided consent for stored samples to be used for future research purposes.

DNA isolation and DNA methylation

DNA was isolated from the prospectively collected and biobanked esophageal biopsies using the DNeasy Blood & Tissue Kit (Qiagen, Germantown, MD), and bisulfite conversion was performed using the EZ DNA Methylation kit (Zymo Research, Irvine, CA). Pretreatment DNA methylation was assayed with the Human MethylationEPIC BeadChip (Illumina, San Diego, CA) and analyzed using ChAMP (33). DNA methylation data were loaded and filtered initially using SampleCutoff = 0.1, which excluded samples with ≥10% of failed CpG sites. Additional quality control measures (as implemented within ChAMP) included removal of probes with a detection P value > 0.01, bead count <3 in at least 5% of the included samples, known single nucleotide polypmorphisms (34), known multihit probes (35), and all chromosome XY probes. After all filtering was performed, the final dataset included 636,218 CpG probes and 88 samples (58 responders and 30 nonresponders). Beta-mixture quantile normalization within ChAMP was performed to adjust for the 2 different assays (Infinium I and II). Single-value decomposition was used to identify potential technical confounders, and adjustments were made for assay plate, chip, and position on chip using Combat (36). These adjusted beta values were used for all analyses.

Statistical analysis

Clinical, endoscopic, and histologic characteristics of the study population were summarized as mean values, SDs, and percentages. To test for an association between the a priori 18 CpG site DNA methylation levels and response to treatment, adjusting for plate, chip, position on the chip, age, sex, and baseline eosinophil count, a logistic regression analysis was conducted. Multiple logistic regression and corresponding area under the receiver operating characteristic curve was computed for replicating CpG sites and for the previous CpG sites previously identified, irrespective of replication in the current sample.

RESULTS Patient characteristics and histologic response status

We analyzed 88 patients with biopsies and DNA methylation data. In this subset of the clinical trial population, the mean age was 37.7 ± 15.5 years, 36% were female, 97% were White, and 72% had at least 1 concomitant atopic condition (Table 1). After treatment, there were 58 histologic responders (66%) and 30 nonresponders (34%) at the <15 eos/hpf threshold. Among responders, peak eosinophil counts decreased from a mean of 69 ± 45 to 2 ± 3 eos/hpf and EREFS decreased from 4.7 ± 1.9 to 1.6 ± 1.5. By contrast, in nonresponders, there was a substantially reduced decrease in eosinophil counts (from 99 ± 69 to 59 ± 40 eos/hpf) and EREFS (5.5 ± 1.5 to 4.4 ± 1.8) (Table 1).

Table 1. - Patient characteristics at baseline and posttreatment, stratified by histologic response status Overall group (n = 88) Histologic nonresponders (n = 30) Histologic responders (n = 58) Age, mean ± SD 37.7 ± 15.5 32.9 ± 13.5 40.2 ± 16.0 Female, n (%) 32 (36) 11 (37) 21 (36) White, n (%) 85 (97) 27 (90) 58 (100) Any atopic condition, n (%) 61 (72) 19 (70) 42 (72) EREFS, mean score ± SD  Baseline   Total 5.0 ± 1.8 5.5 ± 1.5 4.7 ± 1.9   Inflammatory 3.0 ± 1.2 3.4 ± 0.9 2.8 ± 1.2   Fibrostenotic 1.9 ± 1.0 2.0 ± 1.1 1.9 ± 1.0  Posttreatment   Total 2.6 ± 2.1 4.4 ± 1.8 1.6 ± 1.5   Inflammatory 1.2 ± 1.4 2.5 ± 1.3 0.5 ± 0.9   Fibrostenotic 1.3 ± 1.0 1.8 ± 1.1 1.1 ± 0.9 Eosinophil counts, mean eos/hpf ± SD  Baseline 79.3 ± 56.0 99.6 ± 69.4 68.7 ± 44.8  Posttreatment 21.6 ± 35.6 59.1 ± 39.5 2.2 ± 3.4

eos/hpf, eosinophils per high-power field; EREFS, Endoscopic Reference Score.

Validation of DNA methylation sites

Of the 18 CpG sites previously reported to be associated with tCS response, 13 were present on the EPIC Beadchip and met quality control criteria. Of these, 3 CpG sites were associated with responder status after adjusting for covariates (P < 0.012), including sites within UNC5B (cg26152017), ITGA6 (cg01044293), and LRRC8A (cg13962589). All 3 of these sites exhibited a lower proportion methylated in responders when compared with those who did not respond, consistent with the original discovery associations (Table 2). The predictive probability for nonresponse of inclusion of all 3 CpG sites together was strong (area under the receiver operating characteristic curve = 0.79) (Figure 1).

Table 2. - Discovery and replication analysis for differential methylation for treatment response in eosinophilic esophagitis CpG site Chr. Position Discovery (n = 32) Replication (n = 88) Proximal genes Δ βa P value Nonresponders (n = 30) Responders (n = 58) Adjusted logistic Regression P valueb cg23691894 chr10:111765904 ADD3 −0.094 4.6 × 10−6 0.6243 ± 0.0497 0.6336 ± 0.0474 0.6211 cg08810053 chr6:31926073 RDBP; SKIV2L; MIR1236 0.089 1.3 × 10−5 0.5343 ± 0.0427 0.5407 ± 0.0477 0.2515 cg23081781 chr7:127225937 GCC1 −0.092 2.6 × 10−5 0.4797 ± 0.0530 0.4852 ± 0.0555 0.6834 cg18204200 chr14:67999432 PLEKHH1 −0.053 2.7 × 10−5 0.6064 ± 0.0273 0.6158 ± 0.0339 0.8371 cg24407204 chr10:118084109 CCDC172; C10orf96 −0.081 2.8 × 10−5 0.6224 ± 0.0417 0.6209 ± 0.0409 0.9689 cg24839386 chr10:73004577 UNC5B −0.103 3.2 × 10−5 Not assessed cg16320159 chr10:111765674 AAD3 −0.084 4.6 × 10−5 Not assessed cg26152017 chr10:73004559 UNC5B −0.079 4.7 × 10 −5 0.3770 ± 0.0644 0.3313 ± 0.0578 0.0118 cg13868393 chr17:75136250 SEL14L1 −0.028 5.2 × 10−5 0.2560 ± 0.0295 0.2544 ± 0.0311 0.4968 cg08586737 chr7:127225949 GCC1 −0.083 5.4 × 10−5 0.3887 ± 0.0474 0.3963 ± 0.0420 0.4287 cg14074486 chr1:161195356 TOMM40L −0.061 5.8 × 10−5 0.6177 ± 0.0405 0.6121 ± 0.0402 0.4716 cg19121684 chr12:120652844 PXN −0.044 5.9 × 10−5 0.4797 ± 0.0530 0.4852 ± 0.0550 0.5663 cg17022577 chr1:22448980 WTN4 −0.071 6.4 × 10−5 Not assessed g27413385 chr17:7515426 FXR2 −0.084 8.1 × 10−5 Not assessed cg15153068 chr2:2009036 MYT1L 0.046 9.3 × 10−5 Not assessed cg01044293 chr2:173296469 ITGA6 −0.131 9.3 × 10 −5 0.4940 ± 0.0711 0.4210 ± 0.0726 0.0052 cg16575427 chr17:2603795 LIAA0664; CLUH −0.053 9.4 × 10−5 0.5729 ± 0.0273 0.5649 ± 0.0317 0.0732 cg13962589 chr9:131647126 LRRC8A −0.096 9.7 × 10 −5 0.5946 ± 0.0505 0.5496 ± 0.0523 0.0097

aΔ β—defined as the difference between responders and nonresponders.

bAdjustment included age, sex, baseline eosinophil count, plate, chip, and position on the chip.

Figure 1.:

Receiver operating characteristic curve for the 3 identified CpG sites with the area under the curve calculated as 0.7908.

DISCUSSION

Pharmacoepigenetics is emerging as an important tool to inform personalized medicine. In this study, the identification of biomarkers associated with the response to tCS treatment in patients with EoE may enable a treatment strategy tailored to those most likely to respond to tCS, while more efficiently moving to alternative treatments in those likely not to respond. Building on our preliminary data that identified a panel of 18 differently methylated CpG sites associated with tCS response in EoE, we tested for association with response for these same CpG sites in an independent sample to validate the initial findings. While we were not able to validate results from all 18 CpG sites, we corroborated 3 CpG sites with lower methylation in responders than nonresponders, after adjusting for selected baseline clinical features. Given that these same sites exhibited a lower percent methylation in our discovery cohort and are now validated in an independent sample, they have the potential to inform clinical use of topical steroid treatment in EoE. Because additional associations in larger samples are validated, these CpG sites could contribute to a polyepigenetic score that, combined with clinical data, informs treatment strategy.

With the advent of novel treatments for EoE and with multiple trials underway, there is a critical need for individualization of treatment strategies, specifically efficiently assigning patients to the therapies to which they are most likely to respond. This may allow for a decreased need for upper endoscopies to monitor empirically chosen treatments or switching from one ineffective treatment to another option, depending on the complexity of the response outcomes assessed (37). Prior studies assessing clinical predictors of response have shown that nonresponse is associated with esophageal dilation (10,11,14), the so-called “extreme narrow caliber esophagus” (38), increased BMI (39), younger age (14,40), and minoritized racial groups (41). However, results have not yet been replicated, and these patient-based factors are not routinely considered when choosing treatments. Some studies have also assessed immunologic and molecular features, but without consistent or clinically applicable results (15,17,20,42). For example, a pilot study suggested that decreased tissue levels of mast cells and eotaxin-3 were associated with nonresponse (10), but this was not validated (17). Furthermore, a transforming growth factor-β1 single-nucleotide polymorphism (43), a change in FK506-binding protein 5 (FKBP5) mRNA levels (44), differentially expressed genes (45), differentially correlated gene panels (15), and a possible steroid-refractory endotype (46) have all been preliminarily identified but either not confirmed or remain to be validated. In that context, our results are notable because we have now validated 3 methylation sites to be associated with tCS nonresponse.

While the molecular signaling underlying the observed associations between lower percent methylation of UNC5B (cg26152017), ITGA6 (cg01044293), and LRRC8A (cg13962589) cannot be fully elucidated here, hypomethylation of UNC-5 netrin receptor B (UNC5B) has also been associated with poor prognosis in colorectal cancer (47). The UNC5B gene encodes netrin receptors and mediates the effect of netrin-1. Netrin-1 signaling has also been associated with protection of other disease processes, including diabetic kidney disease and dextran sodium sulfate–induced colitis in the mouse model (48,49). Integrin subunit alpha 6 (ITGA6) is a protein-coding gene encoding integrins that interact with extracellular matrix proteins including members of the laminin family. Integrins contribute to processes critical to inflammation and tissue repair (50). Depletion of leucine-rich repeat containing protein-8A (LRRC8A) expression promotes cell apoptosis in the esophagus (51). Epithelial cell apoptosis is associated with epithelial barrier dysfunction (52), which is a key pathogenic feature of EoE (53).

We acknowledge some limitations of this study. First, the parent study was conducted at a single academic center, and participants were primarily adults (though some adolescents were enrolled), male, and White. Despite this, the inclusion criteria limited study subjects to those with a new diagnosis of EoE, so patients should be reflected of this population, not a specialized or more severe referral population. Second, all patients were treated with tCS that were adapted from asthma preparations, so we are not able to comment on the predictive utility of differential epigenetic methylation regarding proton pump inhibitors, diet elimination, biologics, or approved esophageal-specific tCS formulations such as orodispersible budesonide tablets. Third, we focused on histologic response because this was the primary outcome of the parent study. Future investigations could prospectively examine “clinicopathologic” responses based on thresholds for both histologic and clinical responses. Fourth, we did not use the same CpG chip in this validation study as that used in our preliminary study, so this study represents validation of previous study findings, when compared with replication. However, we used a population for independent validation with data and specimens collected during a rigorously conducted clinical trial, with standardized data collection protocols and sample handling and storage. We also used a robust and proven array to measure DNA methylation. These strengths and independent validation lend merit to our findings, though we also acknowledge that the P values for the validated CpG sites are relatively modest.

In conclusion, we validated 3 epigenetic biomarkers (CpG methylation sites) for the prediction of tCS response in patients with EoE in an independent sample, confirming our prior work. While not all previously identified markers were validated, these 3 demonstrated a relatively high predictive probability for nonresponse to tCS treatment and thus hold promise for personalizing therapy in patients with EoE.

CONFLICTS OF INTEREST

Guarantor of the article: Evan S. Dellon, MD, MPH.

Specific author contributions: All authors approved the final draft submitted. E.J.T.: Project conception, study design, data analysis/interpretation, manuscript drafting, critical revision, funding. C.D.L. and T.D.H.: data analysis/interpretation; critical revision. E.S.D.: Project conception, study design, data analysis/interpretation, manuscript drafting, critical revision, funding.

Financial support: This study was supported in part by NIH grants R21 DK122297, K23 DK090073, R01 DK101856, resources from the UNC Center for Gastrointestinal Biology and Disease (P30 DK034987), and by the Center for Public Health Genomics at Wake Forest University School of Medicine.

Potential competing interests: None of the authors have potential conflicts of interest related to this paper. However, E.S.D. is a consultant for Abbott, AbbVie, Adare/Ellodi, Aimmune, Akesobio, Alfasigma, ALK, Allakos, Amgen, Aqilion, Arena/Pfizer, Aslan, AstraZeneca, Avir, Biorasi, Calypso, Celgene/Receptos/BMS, Celldex, Eli Lilly, EsoCap, Eupraxia, Ferring, GSK, Gossamer Bio, Holoclara, Invea, Landos, LucidDx, Morphic, Nextstone Immunology, Nutricia, Parexel/Calyx, Phathom, Regeneron, Revolo, Robarts/Alimentiv, Salix, Sanofi, Shire/Takeda, Target RWE, and Upstream Bio, has received research funding from Adare/Ellodi, Allakos, Arena, AstraZeneca, GSK, Meritage, Miraca, Nutricia, Celgene/Receptos/BMS, Regeneron, Revolo, and Shire/Takeda, and has received education grants from Allakos, Holoclara, and Invea. E.T.J. is a consultant for Regeneron and Jazz Pharmaceuticals and has received research funding from Nutricia.

IRB statement: This study was approved by the University of North Carolina and Wake Forest Institutional Review Boards.

Study Highlights

WHAT IS KNOWN ✓ Swallowed topical corticosteroids (tCS) are a first-line treatment option for patients with eosinophilic esophagitis (EoE). ✓ Few pretreatment clinical or molecular predictors have been identified to direct tCS treatment to those who are most likely to benefit. ✓ We previously identified 18 CpG methylation sites associated with treatment response to tCS.

WHAT IS NEW HERE ✓ Using samples from an independent patient cohort, we assessed baseline (pretreatment) methylation status from the 18 previously identified sites. ✓ Three CpG sites were associated with treatment response, including sites within UNC5B (cg26152017), ITGA6 (cg01044293), and LRRC8A (cg13962589). ✓ Use of these epigenetic markers may be able to guide tCS treatment in EoE. REFERENCES 1. Dellon ES, Liacouras CA, Molina-Infante J, et al. Updated international consensus diagnostic criteria for eosinophilic esophagitis: Proceedings of the AGREE conference. Gastroenterology 2018;155:1022–33.e10. 2. Muir A, Falk GW. Eosinophilic esophagitis: A review. JAMA 2021;326:1310–8. 3. Hirano I, Chan ES, Rank MA, et al. AGA Institute and the Joint Task Force on allergy-immunology practice parameters clinical guidelines for the management of eosinophilic esophagitis. Gastroenterology 2020;158:1776–86. 4. Rank MA, Sharaf RN, Furuta GT, et al. Technical review on the management of eosinophilic esophagitis: A report from the AGA Institute and the Joint Task Force on allergy-immunology practice parameters. Gastroenterology 2020;158:1789–810.e15. 5. Dellon ES, Rothenberg ME, Collins MH, et al. Dupilumab in adults and adolescents with eosinophilic esophagitis. N Engl J Med 2022;387:2317–30. 6. Aceves SS, Dellon ES, Greenhawt M, et al. Clinical guidance for the use of dupilumab in eosinophilic esophagitis: A yardstick. Ann Allergy Asthma Immunol 2023;130(3):371–8. 7. Chang JW, Rubenstein JH, Mellinger JL, et al. Motivations, barriers, and outcomes of patient-reported shared decision making in eosinophilic esophagitis. Dig Dis Sci 2021;66:1808–17. 8. Peery AF, Shaheen NJ, Dellon ES. Practice patterns for the evaluation and treatment of eosinophilic oesophagitis. Aliment Pharmacol Ther 2010;32:1373–82. 9. Eluri S, Iglesia EGA, Massaro M, et al. Practice patterns and adherence to clinical guidelines for diagnosis and management of eosinophilic esophagitis among gastroenterologists. Dis Esophagus 2020;33(7):doaa025. 10. 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;13:452–8. 11. Moawad F, Albert D, Heifert T, et al. Predictors of non-response to topical steroids treatment in eosinophilic esophagitis. Am J Gastroenterol 2013;108(Suppl 1):S14 (abstr 37). 12. Dellon ES, Katzka DA, Collins MH, et al. Budesonide oral suspension improves symptomatic, endoscopic, and histologic parameters compared with placebo in patients with eosinophilic esophagitis. Gastroenterology 2017;152:776–86.e5. 13. Hirano I, Collins MH, Katzka DA, et al. Budesonide oral suspension improves outcomes in patients with eosinophilic esophagitis: Results from a phase 3 trial. Clin Gastroenterol Hepatol 2022;20:525–34.e10. 14. Eluri S, Selitsky SR, Perjar I, et al. Clinical and molecular factors associated with histologic response to topical steroids in patients with eosinophilic esophagitis. Clin Gastroenterol Hepatol 2019;17:1081–8.e2. 15. Dellon ES, Tsai YS, Coffey AR, et al. Pre-treatment differential correlation of gene expression and response to topical steroids in eosinophilic esophagitis. Dis Esophagus 2022;36(4):doac071. 16. Eluri S, Runge TM, Hansen J, et al. Diminishing effectiveness of long-term maintenance topical steroid therapy in PPI non-responsive eosinophilic esophagitis. Clin Transl Gastroenterol 2017;8:e97. 17. Dellon ES, Woosley JT, McGee SJ, et al. Utility of major basic protein, eotaxin-3, and mast cell tryptase staining for prediction of response to topical steroid treatment in eosinophilic esophagitis: Analysis of a randomized, double-blind, double dummy clinical trial. Dis Esoph 2020;33(6):doaa003. 18. Schoepfer AM, Safroneeva E, Bussmann C, et al. Delay in diagnosis of eosinophilic esophagitis increases risk for stricture formation in a time-dependent manner. Gastroenterology 2013;145:1230–6.e2. 19. Dellon ES, Kim HP, Sperry SL, et al. A phenotypic analysis shows that eosinophilic esophagitis is a progressive fibrostenotic disease. Gastrointest Endosc 2014;79:577–85.e4. 20. Jensen ET, Langefeld CD, Zimmerman KD, et al. Epigenetic methylation in Eosinophilic Esophagitis: Molecular ageing and novel biomarkers for treatment response. Clin Exp Allergy 2020;50:1372–80. 21. Dellon ES, Woosley JT, Arrington A, et al. Rapid recurrence of eosinophilic esophagitis activity after successful treatment in the observation phase of a randomized, double-blind, double-dummy trial. Clin Gastroenterol Hepatol 2020;18:1483–92.e2. 22. Dellon ES, Woosley JT, Arrington A, et al. Efficacy of budesonide vs fluticasone for initial treatment of eosinophilic esophagitis in a randomized controlled trial. Gastroenterology 2019;157:65–73.e5. 23. Hirano I, Moy N, Heckman MG, et al. Endoscopic assessment of the oesophageal features of eosinophilic oesophagitis: Validation of a novel classification and grading system. Gut 2013;62:489–95. 24. Dellon ES, Cotton CC, Gebhart JH, et al. Accuracy of the eosinophilic esophagitis endoscopic reference score in diagnosis and determining response to treatment. Clin Gastroenterol Hepatol 2016;14:31–9. 25. Cotton CC, Woosley JT, Moist SE, et al. Determination of a treatment response threshold for the eosinophilic esophagitis endoscopic reference score. Endoscopy 2022;54:635–43. 26. Dellon ES, Fritchie KJ, Rubinas TC, et al. Inter- and intraobserver reliability and validation of a new method for determination of eosinophil counts in patients with esophageal eosinophilia. Dig Dis Sci 2010;55:1940–9. 27. Rusin S, Covey S, Perjar I, et al. Determination of esophageal eosinophil counts and other histologic features of eosinophilic esophagitis by pathology trainees is highly accurate. Hum Pathol 2017;62:50–5. 28. Dellon ES, Veerappan R, Selitsky SR, et al. A gene expression panel is accurate for diagnosis and monitoring treatment of eosinophilic esophagitis in adults. Clin Transl Gastroenterol 2017;8:e74. 29. Dellon ES, Yellore V, Andreatta M, et al. A single biopsy is valid for genetic diagnosis of eosinophilic esophagitis regardless of tissue preservation or location in the esophagus. J Gastrointestin Liver Dis 2015;24:151–7. 30. Wolf WA, Cotton CC, Green DJ, et al. Evaluation of histologic cutpoints for treatment response in eosinophilic esophagitis. J Gastroenterol Hepatol Res 2015;4:1780–7. 31. Reed CC, Wolf WA, Cotton CC, et al. Optimal histologic cutpoints for treatment response in patients with eosinophilic esophagitis: Analysis of data from a prospective cohort study. Clin Gastroenterol Hepatol 2018;16:226–33.e2. 32. Dellon ES, Gupta SK. A conceptual approach to understanding treatment response in eosinophilic esophagitis. Clin Gastroenterol Hepatol 2019;17:2149–60. 33. Tian Y, Morris TJ, Webster AP, et al. ChAMP: Updated methylation analysis pipeline for Illumina BeadChips. Bioinformatics 2017;33:3982–4. 34. Zhou W, Laird PW, Shen H. Comprehensive characterization, annotation and innovative use of Infinium DNA methylation BeadChip probes. Nucleic Acids Res 2017;45:e22. 35. Nordlund J, Bäcklin CL, Wahlberg P, et al. Genome-wide signatures of differential DNA methylation in pediatric acute lymphoblastic leukemia. Genome Biol 2013;14:r105. 36. Johnson WE, Li C, Rabinovic A. Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics 2007;8:118–27. 37. von Arnim U, Biedermann L, Aceves SS, et al. Monitoring patients with eosinophilic esophagitis in routine clinical practice—International expert recommendations. Clin Gastroenterol Hepatol 2022. doi: 10.1016/j.cgh.2022.12.018. [Epub ahead of print December 24, 2022.] 38. Eluri S, Runge TM, Cotton CC, et al. The extremely narrow-caliber esophagus is a treatment-resistant subphenotype of eosinophilic esophagitis. Gastrointest Endosc 2016;83:1142–8. 39. Ketchem CJ, Ocampo AA, Xue Z, et al. Higher body mass index is associated with decreased treatment response to topical steroids in eosinophilic esophagitis. Clin Gastroenterol Hepatol 2023;21:2252–2259.e3. 40. Ketchem CJ, Thakkar KP, Xue A, et al. Older patients with eosinophilic esophagitis have high treatment response to topical steroids. Dig Liver Dis 2022;54:477–82. 41. Ocampo AA, Xue Z, Chang NC, et al. Clinical features and treatment response to topical steroids in ethnic and racial minority patients with eosinophilic esophagitis. Am J Gastroenterol 2022;117(Suppl):e356 (S502; abstr D0207). 42. Aceves SS, King E, Collins MH, et al. Alignment of parent- and child-reported outcomes and histology in eosinophilic esophagitis across multiple CEGIR sites. J Allergy Clin Immunol 2018;142:130–8.e1. 43. Aceves SS, Newbury RO, Chen D, et al. Resolution of remodeling in eosinophilic esophagitis correlates with epithelial response to topical corticosteroids. Allergy 2010;65:109–16. 44. Caldwell JM, Blanchard C, Collins MH, et al. Glucocorticoid-regulated genes in eosinophilic esophagitis: A role for FKBP51. J Allergy Clin Immunol 2010;125:879–88.e8. 45. Butz BK, Wen T, Gleich GJ, et al. Efficacy, dose reduction, and resistance to high-dose fluticasone in patients with eosinophilic esophagitis. Gastroenterology 2014;147:324–33.e5. 46. Shoda T, Wen T, Aceves SS, et al. Eosinophilic oesophagitis endotype classification by molecular, clinical, and histopathological analyses: A cross-sectional study. Lancet Gastroenterol Hepatol 2018;3:477–88. 47. Okazaki S, Ishikawa T, Iida S, et al. Clinical significance of UNC5B expression in colorectal cancer. Int J Oncol 2012;40:209–16. 48. Ranganathan P, Mohamed R, Jayakumar C, et al. Deletion of UNC5B in kidney epithelium exacerbates diabetic nephropathy in mice. Am J Nephrol 2015;41:220–30. 49. Ranganathan P, Jayakumar C, Li DY, et al. UNC5B receptor deletion exacerbates DSS-induced colitis in mice by increasing epithelial cell apoptosis. J Cell Mol Med 2014;18:1290–9. 50. Mezu-Ndubuisi OJ, Maheshwari A. The role of integrins in inflammation and angiogenesis. Pediatr Res 2021;89:1619–26. 51. Konishi T, Shiozaki A, Kosuga T, et al. LRRC8A expression influences growth of esophageal squamous cell carcinoma. Am J Pathol 2019;189:1973–85. 52. Schulzke JD, Bojarski C, Zeissig S, et al. Disrupted barrier function through epithelial cell apoptosis. Ann N Y Acad Sci 2006;1072:288–99. 53. Masterson JC, Biette KA, Hammer JA, et al. Epithelial HIF-1α/claudin-1 axis regulates barrier dysfunction in eosinophilic esophagitis. J Clin Invest 2019;129:3224–35.