1

1 Technical Review on the Management of Eosinophilic Esophagitis: A report from Formatted: Numbering: Continuous 2 the American Gastroenterological Association Institute and the Joint Task Force 3 on Allergy-Immunology Practice Parameters 4 5 Matthew A. Rank MD1 * 6 Ravi N. Sharaf MD, MS2 * 7 Glenn T. Furuta MD3 8 Seema A. Aceves, MD, PhD4 9 Matthew Greenhawt, MD, MSc, MBA5 10 Jonathan M. Spergel, MD, PhD6 11 Yngve T. Falck-Ytter, MD7 * 12 Evan S. Dellon MD MPH8 * 13 14 Collaborators: Bernstein JA9, Dinakar C10 , Golden DBK11, Khan DA12, Lieberman J13, 15 Oppenheimer J14, Shaker M15, Stukus DR16, Wallace DV17, Wang J18 16 17 *These authors have contributed equally to this report. 18 1Division of Allergy, Asthma, and Clinical Immunology, Mayo Clinic, Scottsdale, AZ; 19 2Division of Gastroenterology, Donald and Barbara Zucker School of Medicine at 20 Hofstra/Northwell, Manhasset, NY; 3 Digestive Health Institute, Children’s Hospital 21 Colorado, Gastrointestinal Eosinophilic Diseases Program and University of Colorado 22 School of Medicine, Aurora, CO;4Division of Allergy Immunology Center for Immunity, 23 Infection, and Inflammation, University of California, San Diego Rady Children's 24 Hospital, San Diego, CA; 5 Section of Allergy/Immunology, Children’s Hospital Colorado, 25 University of Colorado School of Medicine, Aurora, CO; 6 Division of Allergy- 26 Immunology, Children’s Hospital of Philadelphia, Perelman School of Medicine at 27 University of Pennsylvania, Philadelphia, PA; 7Division of Gastroenterology and 28 Hepatology, Cleveland VA Medical Center and University Hospitals, Case Western 29 Reserve University School of Medicine, Cleveland, OH; 8 Center for Esophageal 30 Diseases and Swallowing, Division of Gastroenterology and Hepatology, University of 31 North Carolina School of Medicine, Chapel Hill, NC 9 Department of Internal Medicine, 32 University of Cincinnati College of Medicine, Cincinnati, OH; 10 Division of Allergy and 33 Asthma, Stanford University School of Medicine, Palo Alto, CA; 11 Department of Medicine, 34 Johns Hopkins University, Baltimore, MD; 12 Division of Allergy & Immunology, Department 35 of Medicine, University of Texas Southwestern Medical Center, Dallas, TX; 13 Division of 36 Allergy and Immunology, The University of Tennessee Health Science Center, LeBonheur 37 Children’s Hospital, Memphis, TN; 14 Department of Internal Medicine, New Jersey Medical 38 School, Morristown, NJ; 15 Section of Allergy and Immunology, Dartmouth-Hitchcock Medical 39 Center and Dartmouth Geisel School of Medicine, Lebanon and Hanover, NH; 16Division of 40 Allergy and Immunology, Nationwide Children's Hospital and The Ohio State University 41 College of Medicine, Columbus, OH 17 Department of Medicine, Nova Southeastern University, 42 Davie, FL; 18 Division of Allergy and Immunology, Department of Pediatrics, The Elliot and 2

43 Roslyn Jaffe Food Allergy Institute, Icahn School of Medicine at Mount Sinai, Kravis Children’s 44 Hospital, New York NY. 45 46 47 48 Addresses for Correspondence: 49 Chair, Clinical Guidelines Committee, 50 American Gastroenterological Association 51 National Office, 4930 Del Ray Avenue, 52 Bethesda, Maryland 20814. 53 E-mail: ***gastro.org 54 Telephone: (301) 941-2618. 55 56 Joint Task Force on Allergy-Immunology Practice Parameters 57 555 E Wells Street , Suite 1100 58 Milwaukee, WI 53212 59 Email: [email protected] 60 61 62 Manuscript Word Count: 63 References: 116 64 Tables and Figures: 65 eTables and eFigures (in the supplement): 4 and 2 66 67 Keywords: Technical Review; eosinophilic esophagitis; proton pump inhibitor; 68 swallowed corticosteroids; corticosteroids; dietary therapy; elimination diet; elemental 69 diet; targeted elimination diet; biologic therapy; esophageal dilation 70 71 Disclosures and Conflicts of Interest: Conflict of interest disclosure: All members 72 were required to complete disclosure statement. These statements are maintained at 73 the American Gastroenterological Association Institute (AGA) headquarters in 74 Bethesda, Maryland and the Joint Task Force for Allergy-Immunology Practice 75 Parameters in Milwaukee, WI and pertinent disclosure are published with the report. 76 77 Dr. Rank is supported by the Robert E. and Patricia D. Kern Center for the Science of 78 Healthcare Delivery at Mayo Clinic and the Levin Family Foundation. Dr. Sharaf is 79 supported by the National Cancer Institute NCI 1K07CA216326-01A1 and NCI R01 80 1R01CA211723-01A1. Dr. Furuta is supported by the LaCache Chair from Children’s 81 Hospital Colorado 1K24DK100303 (FurutaGT) and Consortium for Gastrointestinal 82 Eosinophilic Researchers (CEGIR). CEGIR (U54 AI117804) is part of the Rare 83 Diseases Clinical Research Network (RDCRN), an initiative of the Office of Rare 84 Diseases Research (ORDR), NCATS, and is funded through collaboration between 85 NIAID, NIDDK, and NCATS and APFED, CURED and EFC. Dr. Aceves is supported by 86 R01, K24 AI, NIAID, and Consortium for Gastrointestinal Eosinophilic Researchers 87 (CEGIR). CEGIR (U54 AI117804) is part of the Rare Diseases Clinical Research 88 Network (RDCRN), an initiative of the Office of Rare Diseases Research (ORDR), 3

89 NCATS, and is funded through collaboration between NIAID, NIDDK, and NCATS and 90 APFED, CURED and EFC. Dr. Greenhawt is supported by grant #5K08HS024599-02 91 from the Agency for Healthcare Quality and Research. Dr. Spergel is supported by the 92 Consortium for Gastrointestinal Eosinophilic Researchers (CEGIR). CEGIR (U54 93 AI117804) is part of the Rare Diseases Clinical Research Network (RDCRN), an 94 initiative of the Office of Rare Diseases Research (ORDR), NCATS, and is funded 95 through collaboration between NIAID, NIDDK, and NCATS and APFED, CURED and 96 EFC and Stuart Starr Chair of Pediatrics. Dr. Dellon is supported by National Institutes 97 of Health grant R01DK101856. 98 99 Acknowledgements: We sincerely thank Kellee Kaulbeck, HBA, MISt for assistance 100 with the medical information search and Stephanie Stanford of the American 101 Gastroenterology Association and Natalie Aumann of the Joint Task Force for Allergy- 102 Immunology Practice Parameters for administrative support. We sincerely thank the 103 following investigators for sharing unpublished data from their studies: Jeff Alexander 104 and Fred Clayton. 105 106 Abbreviations used in this paper: APT, atopy patch test; CI, confidence interval; 107 CRD, component-resolved diagnostic; EGD, esophagogastroduodenoscopy; EoE, 108 eosinophilic esophagitis; GERD, gastroesophageal reflux disease; GRADE, Grading of 109 Recommendations Assessment, Development, and Evaluation; HPF, high-powered 110 field; ITT, intention-to-treat; LR, likelihood ratio; OR, odds ratio; PICO, population, 111 intervention, comparator, and outcome; PPI, proton pump inhibitor; RCT, randomized 112 control trial; RR, risk ratio; sIgE, specific IgE testing in serum; SPT, skin prick testing 113 114 ABSTRACT 115 116 Eosinophilic esophagitis (EoE) is a chronic inflammatory condition of the esophagus. 117 Many new studies have been reported recently that describe EoE management. An 118 expert panel was convened by the American Gastroenterological Association Institute 119 and the Joint Task Force on Allergy-Immunology Practice Parameters to provide a 120 technical report to be used as the basis for an updated clinical guideline. This technical 121 review was developed using the Grading of Recommendations Assessment, 122 Development, and Evaluation (GRADE) framework. Eighteen focused EoE 123 management questions were considered, with 15 answered using the GRADE 124 framework and 3 with a narrative summary. There is moderate certainty in the evidence 125 that topical glucocorticosteroids effectively reduce esophageal eosinophil counts to 126 <15/hpf over a short-term treatment period of 4-12 weeks, but very low certainty about 127 the effects of using topical glucocorticosteroids as maintenance therapy. Multiple 128 dietary strategies may be effective in reducing esophageal eosinophil counts to <15/hpf 129 over a short-term treatment period, with moderate certainty for elemental diets, low 130 certainty for empiric 2,4 and 6 food elimination diets, and very low certainty that allergy- 131 based testing dietary eliminations have a higher failure rate compared to empiric diet 132 elimination. There is very low certainty for the effect of PPIs in patients with esophageal 133 eosinophilia. Although esophageal dilation appears to be relatively safe there is no 134 evidence that it reduces esophageal eosinophil counts. There is very low certainty in 4

135 the effects of multiple other medical treatments for EoE: anti-IL-5 therapy, anti-IL-13 136 therapy, anti-IgE therapy, montelukast, cromolyn, and anti-TNF therapy. 137 138 INTRODUCTION 139 140 Eosinophilic esophagitis (EoE) is a chronic rare inflammatory condition of the 141 esophagus that is estimated to affect 1 in every 2000 people.(1) The incidence of EoE 142 is increasing. EoE can occur in children and adults and is more common in Whites, 143 males, and is associated with other atopic diseases. EoE negatively impacts the quality 144 of life for patients and their families. Medical resource utilization costs in EoE may be 145 significant for some. (2, 3) 146 147 EoE can be characterized by the associated symptoms, visual esophageal endoscopic 148 findings, and histopathology. In adolescents and adults, symptoms often include 149 dysphagia and food impaction, but can be less specific in children, and can include 150 failure to thrive, feeding problems, vomiting, heartburn, and abdominal discomfort. 151 Direct visual inspection of the esophagus in many but not all EoE patients can reveal 152 rings, linear furrows, white plaques or exudates, edema or decreased vascularity, 153 strictures, or luminal narrowing. Histopathology will reveal eosinophils in the 154 esophageal epithelium, which can be defined as a threshold of >15 eosinophils per high 155 power field (hpf). EoE has traditionally been distinguished from gastroesophageal reflux 156 disease (GERD) by the failure of proton pump inhibitor (PPI) treatment to reduce 157 esophageal eosinophilia below a pre-specified threshold. Over the past 10 years, the 158 diagnosis of EoE has been made in a patient who has symptoms of swallowing 159 dysfunction and esophageal eosinophilia that persists despite PPI treatment, and this is 160 the definition that is used as entry criteria for most of the studies presented in this 161 technical report based on previous guidelines. (4-6) However, discerning EoE from 162 GERD remains an area of controversy and active investigation, and the most recent 163 diagnostic criteria for EoE leave the criterion of PPI failure to the clinician (7, 8) since 164 PPI-responsive esophageal eosinophilia now is considered as part of the spectrum of 165 EoE. As such, PPIs are increasingly considered as a treatment rather than as a 166 diagnostic test for EoE as described in a recent consensus document. This 167 recommendation is supported by clinical observations that PPIs resolved EoE-related 168 symptoms and histopathological abnormalities in over 50% of children and adults 169 thought to have EoE. In addition, the biological impact of PPIs to reduce expression of 170 key EoE-related cytokines including eotaxin-3 in vitro and normalize the EoE 171 transcriptome, and the multiple similarities between patients with suspected EoE who 172 do and do not respond to a PPI, together underscore that PPI- responsive esophageal 173 eosinophilia and EoE are potentially disorders in the same pathogenic spectrum. . 174 175 The most common management approaches for EoE are topical glucocorticosteroids, 176 dietary elimination, and esophageal dilation. Many new studies have been published 177 recently. Therefore, the American Gastroenterology Association Institute and the Joint 178 Task Force on Allergy-Immunology Practice Parameters (jointly sponsored by the 179 American College of Allergy, Asthma, and Immunology and the American Academy of 180 Allergy, Asthma, and Immunology) formed a team to provide up-to-date guidance for 5

181 EoE management. This technical review addresses focused clinical questions 182 regarding different therapeutic strategies for managing children and adults with EoE. 183 The results of this technical review were used to inform the development of an 184 accompanying clinical guideline for EoE. 185 186 187 METHODS 188 189 System for Rating the Quality of Evidence 190 191 This technical review and the accompanying guideline were developed using the 192 Grading of Recommendations Assessment, Development, and Evaluation (GRADE) 193 framework.(9) The members of the technical review panel were selected by the AGA 194 Clinical Guidelines Committee and the Joint Task Force on Allergy-Immunology 195 Practice Parameters based on their clinical content and guidelines methodological 196 expertise. Each member underwent a thorough vetting process for potential conflicts of 197 interest. Through an iterative process, and in conjunction with the guideline panel, the 198 participants developed focused clinical questions on the role of specific interventions in 199 the management of EoE. After the focused questions were approved by the 200 organization’s respective leadership groups, the technical review team identified 201 relevant patient-important outcomes, systematically reviewed and summarized the 202 evidence for each outcome across studies, and then rated the quality of the evidence 203 across all outcomes for each clinical question using the GRADE framework. 204 205 Development of Focused Questions 206 207 Using the PICO format, which frames a clinical question by defining a specific 208 Population (P), Intervention (I), Comparator (C), and Outcomes (O), the team developed 209 clinically relevant questions. The PICOs focused on the use of therapeutics in patients 210 with symptomatic EoE. Each of the selected PICO questions was addressed in this 211 review using the GRADE framework except for 2 PICO questions which were 212 addressed using a narrative review format. Studies with children and adults were 213 included. When possible, the interventions were compared to placebo. When only 214 trials compared to another intervention were available, the intervention was presented 215 relative to another intervention (comparator). Appendix Table 1 is a summary display 216 of the 17 PICO questions in this technical report. 217 218 Outcomes 219 220 Potentially relevant patient-important outcomes were considered and rated in terms of 221 importance, as summarized in Appendix Table 2. Through consensus of the expert 222 panel, with no voting necessary during the face-to-face review, and based on precedent 223 literature, failing to achieve histologic remission of < 15 eosinophils/high power field 224 (hpf) was considered critical for decision-making.(10, 11) It was recognized that 225 untreated inflammation can potentially lead to fibro-stenotic disease, (12-15) but also 226 that symptoms do not always correspond with histology. (16-18). Symptoms, changes 6

227 in peak tissue eosinophil levels, and adverse effects were considered important for 228 decision-making. If data on certain outcomes were not available, the a priori plan was 229 to use indirect evidence to guide decision making if additional data were not provided 230 after contacting the investigators. 231 232 Outcomes that are reported in the evidence profiles are those that were found in the 233 literature. Several outcomes that were rated as important by the expert panel are not 234 reported in the evidence profiles because they were not assessed in the included 235 literature. Symptoms were reported using many different scales. Validated EoE 236 symptom questionnaires were not available when most of the studies were performed. 237 Therefore, symptom severity was an outcome that could not be synthesized in a 238 summary estimate due to this heterogeneity in reporting. Similarly, not all studies 239 utilized a validated endoscopy score, and endoscopic outcomes could not be 240 synthesized. Finally, a key decision in forming the estimate of the effect for 241 observational studies lacking a contemporaneous control group was to use the placebo 242 control arm rate for failing to achieve histologic remission from topical corticosteroid 243 studies. The expert panel was in consensus, with no voting needed, that the 86.7% 244 estimate for failing to achieve histologic remission (> 15 eosinophils/hpf) in the placebo 245 arm during a study period of 8 weeks was reasonable based on the overall information 246 available in the literature. (Alexander 2012,(19) Butz 2014,(20) Dellon 2017,(21) Dohil 247 2010,(22) Gupta 2015,(23) Konikoff 2006,(24) Miehlke 2016,(25) Straumann 2010(26)) 248 249 Systematic Review Process 250 251 A common approach to study selection was used for each question. For all PICOs, we 252 first considered high quality systematic reviews for evidence synthesis, particularly 253 those that synthesized data from RCTs. If systematic reviews of RCTs were not 254 available, we then looked to individual RCTs and generated summary estimates as 255 needed. Systematic reviews of observational studies, and in particular, single arm 256 cohort/observational studies, were considered as the least-preferred option to inform the 257 evidence, with rates pooled when possible. Case series with < 5 cases and case reports 258 were excluded, unless no other evidence for the question was available. Systematic 259 reviews that were missing recent trial data were updated and re-analyzed rather than 260 creating a de novo systematic review. When well-done systematic reviews were 261 unavailable, we searched for primary articles using a preliminary search strategy. Next, 262 preliminary evidence profiles were constructed using GRADEPRO 263 (https://gradepro.org/), and were reviewed iteratively with the clinical experts (SAA, 264 GTF, MG, JMS, and ESD), where feedback was provided about missing studies, 265 missing data, and preliminary evidence ratings. 266 267 An additional, final systematic literature search was performed after the preliminary 268 evidence profiles were constructed and reviewed with the expert panel to ensure 269 completeness. Details of the search strategy are reported in the Appendix Table 3. 270 We conducted an electronic search using MEDLINE, EMBASE, and the Cochrane 271 Library until May 13, 2018. A research librarian (KK) developed a single search strategy 272 for MEDLINE and then adapted to EMBASE and Cochrane. The search strategy was 7

273 iteratively refined to maximize sensitivity, working directly with the clinical experts. The 274 search excluded letters, commentaries, editorials, notes, conference abstracts, and 275 non-human studies. We only searched for clinical trials in the electronic literature 276 search for all PICO questions except for the dietary interventions where observational 277 studies were considered, a decision made by the expert panel after considering the 278 preliminary evidence profiles. We searched the WHO clinical trial registry to identify 279 additional studies (http://apps.who.int/trialsearch/). Titles and abstracts were reviewed in 280 duplicate by two authors (MR and RS). One methodologist (RS or MR) extracted data 281 from eligible reports and a second methodologist (RS or MR) evaluated the accuracy of 282 the data extraction. We contacted authors when key data were missing, first by 283 attempting to reach the corresponding author by email and then by trying a 2nd author 284 from the article if no response. Disagreements were resolved by discussion with a third 285 methodologist (YFY). 286 287 Statistical Analysis 288 289 Pooled risk ratios with 95% confidence intervals (CI) were calculated when possible, 290 using RevMan v5.3 (Cochrane Collaboration, Copenhagen, Denmark), or Open 291 Meta[analyst] (Brown University, Providence, RI), particularly when single arm rates 292 were pooled. In RevMan, analyses were performed using a random-effects model. In 293 OpenMeta[analyst], we used binary random effects using the DerSimonian Laird 294 method. Statistical heterogeneity was assessed using the I2 statistic. Publication bias 295 was assessed using funnel plots when possible. GRADEpro software was used to 296 construct the evidence profiles and calculate the absolute effects. When historical 297 controls were used, risk ratios (RRs) were presented and the resulting absolute effects 298 were informed by applying the baseline risk from the untreated control arms from steroid 299 RCTs to the RR. It is important to note that RR refers, in this technical report, to the risk 300 of not achieving histologic remission in the treatment versus a comparator. 301 302 303 RESULTS 304 305 306 #1 Should proton pump inhibitors be used in patients with esophageal 307 eosinophilia? 308 309 Evidence Summary: We identified 23 observational studies which reported that 58.3% 310 (unweighted) of subjects on PPI failed to achieve histopathologic remission (eosinophils 311 <15/hpf) compared to 86.7% (unweighted) of a placebo comparison group. 312 313 Quality of Evidence: The certainty in the effect estimate was very low. The certainty in 314 the estimate was downgraded for inconsistency. 315 316 Discussion: This question is related to patients with esophageal eosinophilia, who, 317 depending on the study and inclusion criteria, may be different than patients with EoE. 318 It is important to note that this is an indirect comparison because participants in the 8

319 topical corticosteroid studies had failed PPI treatment. Understanding PPI response in 320 EoE remains an active area of investigation. The inconsistency seen in the point 321 estimates for histologic response was not clearly explained by any specific criteria (e.g. 322 pediatrics versus adult or inclusion/exclusion criteria). There were 2 RCTs identified 323 that compared PPI to topical corticosteroid, and found similar rates of histologic 324 remission (see Appendix Table 4). 325 PPI compared to placebo for eosinophilic esophagitis

Bibliography: Garrean 2009,(27) Peterson 2010,(28) Molina-Infante 2011,(29) Abe 2011,(30) Fujiwara 2012,(31) Francis 2012,(32) Vazquez-Elizondo 2013,(33) Moawad 2013,(34) Lee 2013(35) Dellon 2013,(36) Mangla 2014(37), Molina 2014,(38) van Rhijn 2014,(39) Gomez-Torrijos 2016,(40) Jiao 2016,(41) Savarino 2017,(42) Philpott 2016,(43) Sayej 2009,(44) Dranove 2010,(45) Schroeder 2013,(46) Rea 2013,(47) Dhaliwal 2014,(48) Gutierrez-Junquera 2016(49) Outcomes and № of Certainty of Relative Anticipated absolute Follow-up participants the effect effects (studies) evidence (95% CI) (GRADE) Risk with Risk placebo difference with PPI

Not achieving 1051 ⨁◯◯◯ RR 0.66 867 per 295 fewer histologic (23 VERY (0.61 to 1,000 per 1,000 remission (<15 observational LOW b,c 0.72) d (338 eos/hpf); studies) a fewer to follow up: mean 8 243 weeks fewer)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

326 Explanations 327 a. Included 2 RCTs of PPI vs topical steroids 328 b. I-squared = 81%; very inconsistent results in absolute terms 329 c. Patients are different than for other interventions where PPI responders were 330 excluded 331 d. Used historical control cohort of placebo arm of topical steroid studies 332 333 9

334 #2 Should topical glucocorticosteroids be used in patients with EoE? 335 336 Evidence Summary: Eight double-blind placebo-controlled RCTs were identified. 337 Summary estimates indicate that 35.1% of patients treated with glucocorticosteroids 338 failed to achieve histologic remission compared to 86.7% of patients treated with 339 placebo, leading to a RR of 0.39, (95% CI 0.26-0.58). Adverse events were experienced 340 by 43% of patients in the topical glucocorticosteroid group compared to 36% of those 341 exposed to placebo, with a risk ratio of 1 (95% CI 0.85-1.19). 342 343 Quality of Evidence: The certainty in the effect estimates was moderate for the outcome 344 of histologic response. We downgraded for inconsistency due to heterogeneity 345 (I2=77%). The certainty in the effect estimates was low for the outcome of adverse 346 events. We rated down for indirectness given heterogeneity in how adverse events 347 were defined and for imprecision given that the risk ratio crossed 1. 348 349 Discussion: RCTs were excluded if they did not have an explicit glucocorticosteroid vs 350 placebo comparison,(28, 50-53) and if they did not include budesonide or fluticasone in 351 the treatment group (54). Six meta-analyses were reviewed and excluded because they 352 did not include the most recent RCTs published in the field, or included studies in 353 addition to a placebo/glucocorticosteroid comparison. 354 355 After discussion amongst the expert panel, the following decisions were made regarding 356 how to pool the data: a single pooled estimate is presented despite differences in type 357 of glucocorticosteroid, delivery mechanism, dosages, patient population 358 (adult/pediatric), and manner of outcome reporting (peak vs mean counts). Notably, 359 sensitivity analyses isolating these individual groups did not alter findings significantly, 360 lending credence to the decision to pool topical glucocorticosteroid data. Most trials 361 required a failed PPI treatment trial prior to enrolling subjects, or excluded patients with 362 GERD. 363 364 Similar categories of data were reported across the 8 included RCTs on 3 outcomes: 365 Histologic response (defined as any eosinophils <15/hpf), symptomatic response, and 366 adverse events. For histologic response: a) data are presented as failure to achieve 367 histologic response, so a RR of <1 means that patients in a given arm are less likely to 368 fail to achieve histologic response, and b) We approximated intention-to-treat (ITT) 369 estimates if not reported, by examining the CONSORT diagram and accounting for 370 dropout. All participants who dropped out in any study arm were categorized as failing 371 to achieve histologic remission. 372 373 For adverse events, there was a variable definition of adverse events (ranging from 374 general (fever/fatigue) to skin/respiratory/GI/endocrine disorders/infections, to those that 375 needed drug-discontinuation). Numbers for adverse events were taken from per 376 protocol analyses (when possible). Potential adverse events have been summarized by 377 Philpott et al. and include local infections-candida and viral, adrenal suppression, 378 diminished growth and fractures.(55) 379 10

380 381 Topical Glucocortiosteroids compared to placebo for eosinophilic esophagitis Bibliography: Alexander 2012,(19) Butz 2014,(20) Dellon 2017,(21) Dohil 2010,(22) Gupta 2015,(23) Konikoff 2006,(24) Miehlke 2016,(25) Straumann 2010(26)

Outcomes and Follow- № of Certainty of Relative Anticipated absolute up participants the effect effects (studies) evidence (95% Risk with Risk (GRADE) CI) placebo difference with Topical Steroids

Not achieving histologic 437 ⨁⨁⨁◯ RR 0.39 880 per 537 fewer remission (<15 eos/hpf); (8 RCTs) a MODERATE (0.26 to 1,000 per 1,000 follow up: mean 8 b,c,d,e 0.58) (from 369 weeks fewer to 651 fewer)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; OR: Odds ratio

382 Explanations 383 a. There was variability in whether histologic response was reported as Mean vs Peak 384 Eosinophil levels. All included trials reported Peaks with the exception of: a) Straumann 385 2010, which reported mean counts, b) Alexander 2012, which reported mean peaks 386 (personal communication), and c) Miehlke reported eosinophil density which we 387 correlated to eosinophil counts. Removing these individual trials in a sensitivity analysis 388 did not significantly alter the summary estimates (0.59 [0.46, 0.76]. 389 b. Few studies reported ITT analyses and dropout was not adequately accounted for. 390 We accounted for this by looking at the CONSORT diagram and reporting ITT results, 391 assuming that outcomes would favor the control. 392 c. I2= 77%. 11

393 d. Dellon 2017 might have had a more severe baseline patient population, accounting 394 for decreased response to glucocorticosteroids when compared to other studies. We did 395 not however rate down for indirectness 396 e. Despite the fact that there are < 300 events (which can indicate suboptimal 397 information size), we did not rate down for imprecision. We also felt that because 398 inconsistency and imprecision are related concepts, a single down-grade for 399 inconsistency was sufficient. We assessed clinical behavior at the extremes of the 400 confidence interval, and judged that behavior would not change. 401 402 403 #3 Should systemic glucocorticosteroids be used in patients with EoE? 404 405 Evidence Summary: We identified 1 RCT that compared prednisone to fluticasone in 406 the treatment of EoE in children. We reported outcomes of “lack of histologic response” 407 defined as failure to achieve <15 eosinophils per hpf, and adverse events, a composite 408 endpoint defined in the footnotes. 11/40 (28%) patients in the prednisone arm vs 14/40 409 (35%) patients in the fluticasone arm failed to achieve histologic response, for a RR 410 0.79 (95% CI 0.41 to 1.52). 16/40 (40%) patients in the prednisone arm compared to 411 6/40 (15%) in the fluticasone arm experienced adverse events for a RR of 2.67 (1.16 to 412 6.11). 413 414 Quality of Evidence: The certainty in the estimates was moderate. Both outcomes were 415 rated down for imprecision; the RR for clinical response had a confidence interval which 416 crossed 1 and adverse events had few events. 417 418 Discussion: A single RCT comparing systemic and topical glucocorticosteroids 419 suggests similar efficacy but a higher rate of adverse events for patients receiving 420 systemic glucocorticosteroid. Systemic adverse events were reported as a composite 421 endpoint, defined in the study as hyperphagia, weight gain, and/or cushingoid features. 422 Other potential adverse effects that have a longer potential time frame to develop such 423 as effects on bone health, immunity, cataract formation, glucose levels, and blood 424 pressure were not measured. In the prednisone group, 16/40 (40%) experienced 425 systemic adverse events; 3 of these 16 exited the study before week 4 and were 426 transitioned to the fluticasone group (outside the protocol). In the fluticasone group, 6/40 427 (15%) patients experienced esophageal candida overgrowth, though none did in the 428 prednisone arm; all of these candida esophageal patients were free of the presenting 429 symptoms by week 4. This single study was conducted in children and may not be 430 applicable to adults. 431 432 433 12

Systemic glucocorticosteroid compared to topical glucocorticosteroid for eosinophilic esophagitis Bibliography: Schaefer 2008(51)

Outcomes and Follow- № of Certainty of Relative Anticipated up participants the effect absolute effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference topical with steroid systemic steroid

Not achieving histologic 80 ⨁⨁⨁◯ RR 0.79 350 per 73 fewer remission (<15 eos/hpf); (1 RCT) MODERATE (0.41 to 1,000 d per 1,000 follow up: mean 8 weeks a,b,c 1.52) (207 fewer to 182 more)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

434 a. Despite a lack of blinding, the study was not downgraded for risk of bias. 435 b. We did not rate down for indirectness despite the fact that this trial consists of a 436 pediatric population. 437 c. Risk Ratio crosses 1 438 d. In order to estimate an ITT analysis, we changed the denominator to the number of 439 subjects who were originally randomized to each group (40 per group). The numerator 440 denotes the number of individuals who achieved at “mild’ or normal histologic grade (per 441 Schaefer paper) at 4 weeks, which correlates to a cutoff of <15Eos/hpf. 442 443 444 445 #4 Should an elemental diet be used in patients with EoE? 446 447 Evidence Summary: We identified 6 observational studies which reported that 6.4% of 448 subjects on elemental diet failed to achieve histopathologic remission (eosinophils 449 <15/hpf) compared to 86.7% in a placebo comparison group (taken as a historical 13

450 comparison group from swallowed topical steroid data) for a RR of 0.07, (95% CI 0.05- 451 0.12). 452 453 Quality of Evidence: The certainty in the effect estimate was moderate. The certainty in 454 the estimate was rated up for anticipated large effect. 455 456 Discussion: There were differences in the effect estimates when grouping children and 457 adult studies, with adult studies having a lower proportion of study participants 458 achieving histologic remission. This comparison is limited by use of a historical 459 comparison group comprised of placebo treated patients in topical steroid studies, 460 which is an indirect but permissible method under GRADE to handle such situations 461 where only single arm observational studies exist. Symptom response was reported for 462 4 studies but could not be synthesized due to considerable differences in the way 463 symptoms were reported. Of the 6 studies used for the efficacy assessment, 3 464 specifically measured nutritional status and 1 measured overall quality of life. Difficulty 465 adhering to an elemental diet was raised as an important consideration for this 466 intervention. Potential harms of this intervention were raised by the expert panel and 467 include the interruption of developmental progress of eating for children, the potential 468 need for gastrostomy tube placement, and the risks associated with repeated 469 endoscopies needed when food is ultimately re-introduced. Risk of developing IgE- 470 mediated food allergy following a period of food elimination has not been described in 471 EoE, but has been described in case reports of children with atopic dermatitis (56). 472 Risk of prolonged peanut avoidance versus early introduction of peanut in the first year 473 of life has been shown as a factor influencing peanut allergy development in children 474 with either severe eczema and/or known egg allergy,(57) but has not been described in 475 EoE and it is unclear how such data would therefore apply. Consultation with an 476 allergist would be recommended in this situation to manage potential competing risks 477 and harms with avoidance diets that would prolong introduction of foods such as peanut 478 (and possibly egg) in children in their first year of life, and potentially place them at risk 479 for developing IgE mediated food allergy. Finally, the expert panel noted that the 480 consideration of an elemental diet would be made in the context of other management 481 options, including other dietary management options such as empiric food elimination 482 (e.g. six food elimination diet) or testing-based elimination diet and with careful 483 consideration for the age of the patient, potential detrimental effects of widespread food 484 elimination, and patient preferences. 485 14

Elemental Diets compared to placebo for the management of Eosinophilic Esophagitis Bibliography: Kelly 1995a,(58) Licarous 2005b,(59) Henderson 2012,(60) Peterson 2013C,(61) Leung 2015,(62) Warners 2017(63) Outcomes and Follow- № of Certainty of Relative Anticipated up participants the effect absolute effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with Elemental Diets

Not achieving histologic 431 ⨁⨁⨁◯ RR 0.07 880 per 807 fewer remission (eos<15/hpf) (6 MODERATE (0.05 to 1,000 per 1,000 follow up: mean 8 observational d 0.12) (824 fewer weeks studies) to 763 fewer)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

486 Explanations 487 a. 5 subjects excluded for noncompliance to diet; 2 subjects excluded for not getting 488 follow-up endoscopy 489 b. 8 subjects excluded for noncompliance 490 c. Histologic remission defined as < 10 eos/hpf; 11 who started diet dropped out 491 d. Upgraded for very large effect size 492 493 #5 Should an empiric food elimination diet be used in patients with EoE? 494 495 A. Should an empiric 6 food elimination diet be used in patients with EoE? 496 497 Evidence Summary: We identified 10 single armed observational studies which 498 reported that 32.1% (unweighted) of subjects on an empiric elimination diet failed to 499 achieve histopathologic remission (eosinophils <15/hpf) compared to 86.7% of a 15

500 placebo comparison group (taken as a historical comparison group from swallowed 501 topical steroid data), for a RR=0.38, (95% CI 0.32-0.43). 502 503 Quality of Evidence: The certainty in the effect estimate was low due to non- 504 comparative single arm study designs. 505 506 Discussion: Symptom response was reported for 3 studies but could not be 507 synthesized due to considerable differences in the way symptoms were reported. Of 508 the 10 studies used for the efficacy assessment, 2 specifically measured nutritional 509 status and 1 formally measured overall quality of life. Difficulty adhering to an empiric 510 diet where 6 foods were eliminated was raised as an important consideration for this 511 intervention. Empiric diet approaches with fewer foods may improve adherence to 512 dietary avoidance and be associated with fewer endoscopies required to identify food 513 triggers. Potential harms of this intervention were raised by the expert panel and 514 include the effect on nutrition and the risks associated with repeated endoscopies 515 needed when food is ultimately re-introduced. Finally, the expert panel noted that the 516 consideration of an empiric diet would be made in the context of other management 517 options, including other dietary management options such as elemental or testing-based 518 elimination diets. 519 520 SFED compared to placebo for the management of Eosinophilic Esophagitis

Bibliography: Kagalwalla 2006a,(64) Gonsalves 2012b,(65) Henderson 2012c,(66) Lucendo 2013d,(67) Colson 2014e,(68) Rodriguez-Sanchez 2014f,(69) Philpott 2016g,(70) Molina-Infante 2018h,(71) Reed 2017i,(72) Homan 2015(73) Outcomes № of Certainty Relative Anticipated participants of the effect absolute effects (studies) evidence (95% Risk Risk Follow-up (GRADE) CI) with difference placebo with SFED

Failure to achieve 633 ⨁⨁◯◯ RR 0.38 880 per 550 fewer histologic remission (9 LOW j (0.32 to 1000 per 1000 (Proportion) observational 0.43) (600 fewer assessed with: studies) to 500 Esophageal eosinphils < fewer) 15/hpf follow up: mean 6 weeks 16

SFED compared to placebo for the management of Eosinophilic Esophagitis Bibliography: Kagalwalla 2006a,(64) Gonsalves 2012b,(65) Henderson 2012c,(66) Lucendo 2013d,(67) Colson 2014e,(68) Rodriguez-Sanchez 2014f,(69) Philpott 2016g,(70) Molina-Infante 2018h,(71) Reed 2017i,(72) Homan 2015(73) Outcomes № of Certainty Relative Anticipated participants of the effect absolute effects (studies) evidence (95% Risk Risk Follow-up (GRADE) CI) with difference placebo with SFED

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

521 Explanations 522 a. Used < 10 eos/hpf for definition of remission. Studies with a cutoff of < 15 eos/hpf 523 will likely underestimate the response rate within the pooled results and therefore 524 strengthen the assumed estimate of effect. 525 b. Measured SF-36 526 c. 15/26 subjects did SFED + foods with + SPT/APT 527 d. Also eliminated rice + corn 528 e. Also eliminated foods with + SPT/APT 529 f. Excluded subjects with + IgE tests prior to enrollment. The response rate is lower 530 and therefore likely underestimates the effect estimate in the pooled results. 531 g. Only subjects with < 5 eos/hpf (“complete” remission) 532 h. Combined clinical and histopathological remission; estimated for SFED based on 2- 533 4-6 FED step-up protocol 534 i. Did not include Wolf 2014 in analysis to avoid duplicate subjects 535 j. When compared to historical controls, a very large effect estimate is likely. 536 However, the evidence certainty was not rated up due to concerns of possible residual 537 confounding and/or indirectness. 538 539 540 B. Should an empiric 4 food elimination diet be used in patients with EoE? 541 542 Evidence Summary: We identified 3 single armed studies which reported that 43.1% 543 (unweighted) of subjects on an empiric elimination diet failed to achieve histopathologic 544 remission (eosinophils <15/hpf) compared to 86.7% (unweighted) of a placebo 17

545 comparison group (taken as a historical comparison group from swallowed topical 546 steroid data), for a RR=0.46, (95% CI 0.42-0.57). 547 548 Quality of Evidence: The certainty in the effect estimate was low due to non- 549 comparative single arm study designs. 550 551 Discussion: The 6 food elimination diet estimate for not achieving histologic remission 552 was slightly lower but similar (32% compared to 43%) than for 4 food elimination diet. 553 Similar to the 6 food elimination diet, potential harms of this intervention were raised by 554 the expert panel and include the effect on nutrition and the risks associated with 555 repeated endoscopies needed when food is ultimately re-introduced. 556 557 4 Food empiric elimination diet compared to placebo for eosinophilic esophagitis

Bibliography: Molina-Infante 2014,(38) Kagalwalla 2017,(74) Molina-Infante 2017a (71)

Outcomes and Follow- № of Certainty Relative Anticipated absolute up participants of the effect effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with 4 Food empiric elimination diet

Not achieving histologic 426 ⨁⨁◯◯ RR 0.49 880 per 449 fewer remission (No remission ) (3 LOW (0.42 to 1,000b per 1,000 assessed with: eos < 15/ observational 0.57) (510 fewer hpf studies) to 378 follow up: 6 weeks fewer)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio 18

558 Explanations

559 a. Pediatric study 560 b. Placebo group responses from topical steroid trials

561 562 C. Should an empiric 2 food elimination diet be used in patients with EoE? 563 564 Evidence Summary: We identified 2 single armed studies which reported that 57.9% 565 (unweighted) of subjects on an empiric elimination diet failed to achieve histopathologic 566 remission (eosinophils <15/hpf) compared to 86.7% (unweighted) of a placebo 567 comparison group (taken as a historical comparison group from swallowed topical 568 steroid data) for a RR=0.66, (95% CI 0.57-0.77). 569 570 Quality of Evidence: The certainty in the effect estimate was very low due to non- 571 comparative single arm study designs and was further rated down for imprecision due to 572 low information size. 573 574 Discussion: The 6 and 4 food elimination diet estimates for not achieving histologic 575 remission were slightly lower than for 2 food elimination (32% and 43% compared to 576 58%). In the Molina-Infante study, the 2 foods eliminated were milk and wheat. In the 577 Reed study, the 2 foods were milk and soy, and the participants had previously been 578 treated with a combination of topical steroids and 2-food elimination in the prior 3 579 months. Potential harms of this intervention were raised by the expert panel and 580 include the effect on nutrition and the risks associated with repeated endoscopies 581 needed when food is ultimately re-introduced although fewer for an empiric 2 food 582 elimination diet than a 4 or 6 food elimination diet. Finally, the expert panel noted that 583 the consideration of an empiric diet would be made in the context of other management 584 options, including other dietary management options such as elemental, other empiric 585 elimination strategies, and testing-based dietary elimination. 586 19

2 food elimination diet compared to placebo for eosinophilic esophagitis Bibliography: Molina-Infante 2017a,(71) Reed 2018b (75)

Outcomes and Follow- № of Certainty Relative Anticipated absolute up participants of the effect effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with 2 food elimination diet

Not achieving histologic 311 ⨁◯◯◯ RR 0.66 880 per 299 fewer remission (Remission) (2 VERY (0.57 to 1,000c per 1,000 assessed with: Eos < observational LOW 0.77) (378 fewer 15/hpf studies) to 202 follow up: 6 weeks fewer)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

587 a. 2FED=milk + wheat 588 b. 2FED=milk + soy 589 c. Placebo group is from topical steroid trials

590 591 D. Should an empiric single food elimination diet be used in patients with 592 EoE? 593 594 Evidence Summary: We identified 2 single armed studies which reported that 45.9% 595 (unweighted) of subjects on a single food empiric elimination diet failed to achieve 596 histopathologic remission (eosinophils <15/hpf) compared to 86.7% (unweighted) of a 597 placebo comparison group (taken as a historical comparison group from swallowed 598 topical steroid data). 599 600 Quality of Evidence: The certainty in the effect estimate was very low due to non- 601 comparative single arm study designs and was further rated down for imprecision due to 602 low information size. 20

603 604 Discussion: The 6 and 4 food elimination diet estimates for not achieving histologic 605 remission were slightly lower than for single food elimination for 2 food elimination (32% 606 and 43% compared to 46%) but was lower than for 2 food elimination (58%). The 607 higher rates of remission with single food (milk) compared to 2-food (both of which 608 included milk) are not easily explained based on study design or patient characteristics 609 and are very uncertain based on the assessment of quality of the evidence. While the 610 risks with a single food elimination strategy are lower compared to 2,4, or 6 food 611 elimination, similar potential harms of this intervention were raised by the expert panel 612 which include the effect on nutrition and the risk associated with endoscopy (though 613 only 1 follow endoscopy because only 1 food eliminated) Finally, the expert panel 614 noted that the consideration of an empiric diet would be made in the context of other 615 management options, including other dietary management options such as elemental, 616 other empiric elimination strategies, and testing-based dietary elimination. 617 Single food elimination compared to placebo for eosinophilic esophagitis

Bibliography: Kagalwalla 2012a,(76) Kruszewski 2016b (77)

Outcomes № of Certainty Relative Anticipated absolute participants of the effect effects (studies) evidence (95% Risk Risk Follow-up (GRADE) CI) with difference placebo with Single food elimination

Not achieving histologic 203 ⨁◯◯◯ RR 0.52 880 per 422 fewer remission (Remission) (2 VERY (0.37 to 1,000d per 1,000 assessed with: eos < observational LOWc 0.74) (554 fewer 15/hpf studies) to 229 follow up: 6 weeks fewer)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

618 Explanations 21

619 a. Milk 620 b. Milk, 6 dropped out and were considered treatment failures for this analysis 621 c. Low information size 622 d. Placebo data are from topical steroid studies

623 624 #6 Should allergy-based testing be used for the purpose of identifying food 625 triggers in patients with EoE? 626 627 Evidence Summary: We identified 12 single armed studies which reported that 49.2% 628 (unweighted) of subjects on a testing-based elimination diet failed to achieve 629 histopathologic remission (eosinophils <15/hpf) compared to 86.7% (unweighted) of a 630 placebo comparison group (taken as a historical comparison group from swallowed 631 topical steroid data). 632 633 Quality of Evidence: The certainty in the effect estimate was very low due to non- 634 comparative single arm study designs. 635 636 Discussion: Inconsistency was noted and thought to be most likely related to the 637 different testing approaches that were used to inform the dietary elimination. Different 638 studies used different testing techniques, or combinations of techniques, including skin 639 prick testing, serum IgE testing, or patch testing. Some studies used all 3 methods to 640 select the dietary intervention. We performed a sensitivity analysis that found 41% 641 (95% CI: 18% to 64%) failed to achieve remission in studies in which patch testing was 642 used and 61% (95% CI: 38%-83%) in studies not using patch testing. Thus, there were 643 more favorable outcomes in studies in which patch testing was performed, but the 644 outcomes weren’t clearly better (considerable CI overlap) and there is very low certainty 645 in this comparative effect estimate. Symptom response was reported for 4 studies but 646 could not be synthesized due to considerable differences in the way symptoms were 647 reported. Of the 10 studies used for the efficacy assessment, 1 specifically measured 648 nutritional status and none formally measured overall quality of life. Difficulty adhering 649 to an elimination diet was raised as an important consideration for this intervention. 650 Potential harms of this intervention were raised by the expert panel and include the 651 effect on nutrition and the risks associated with repeated endoscopies needed when 652 food is ultimately re-introduced. The expert panel noted that the consideration of a 653 testing based diet would be made in the context of other management options, including 654 other dietary management options such as elemental or empiric dietary elimination. 655 Finally, the expert panel discussed the potential role of aeroallergen testing and 656 treatment in EoE. There is emerging evidence that aeroallergens may be important 657 triggers for EoE. There are currently only very small case series reporting aeroallergen 658 immunotherapy in patients with EoE. 659 660 661 22

Allergy-Based Elimination Diets compared to placebo for management of EoE Bibliography: Licarous 2005a,(59) Qaugletta 2007b,(78) Rizo Pascual 2011b,(79) Henderson 2012,(66) Molina-Infante 2012c,(80) Spergel 2012,(81) Al-Hussaini 2013,(82) Rodriguez-Sanchez 2014,(69) Homan 2015,(73) Syrigou 2015d,(83) Van Rhijn 2015e,(84) Constatine 2017,(85) Reed 2017(72) Outcomes and Follow- № of Certainty Relative Anticipated absolute up participants of the effect effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with Allergy- Based Elimination Diets

Not achieving histologic 830 ⨁◯◯◯ RR 0.57 880 per 373 fewer response (11 VERY (0.33 to 1,000g per 1,000 assessed with: < 15 observational LOW f 0.73) (581 fewer eos/hpf studies) to 234 fewer) Follow up: 8 weeks

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

662 Explanations 663 664 a. Skin prick test (SPT) + atopy patch test (APT), some had previously failed 665 pharmacologic therapy 666 b. Remission defined as < 10 eos/hpf 667 c. SPT + APT + PPT 668 d. SPT + specific IgE testing (sIgE) + APT 669 e. Component-resolved diagnostics (CRD) 670 f. Very inconsistent results 671 g. Placebo group responses from topical steroid trials 672 673 #7 Should maintenance therapy be recommended in patients with EoE? 674 23

675 Evidence Summary: There is 1 RCT of continuing therapy compared to placebo for 676 patients who had achieved clinical and histologic remission. The risk ratio was 0.70 677 (95% CI 0.38-1.30) for failing to maintain histologic remission, defined for that study as 678 eosinophils < 20/hpf. 679 680 Quality of Evidence: The certainty in the effect estimate was very low. The certainty in 681 the estimate was rated down for indirectness as the intervention used a delivery 682 mechanism and dose of topical corticosteroid that is different than most previously 683 reported corticosteroid studies. The certainty was also downgraded twice for very 684 serious imprecision due to very low information size. 685 686 Discussion: A single very small RCT of low dose topical glucocorticosteroid (0.25 mg 687 budesonide twice daily) failed to show or to exclude a beneficial effect on maintaining 688 remission in patient who had previously achieved it, using an absolute threshold of <20 689 eos/hpf. However, no patient in the placebo group met the definition of complete 690 response at one year (<5 eos) compared to 36% of the active arm, which was a strong 691 trend (p=0.06). Similarly, the absolute eosinophil counts were significantly lower in the 692 treatment arm compared to placebo (32 eos/hpf vs 65 eos/hpf; p=0.02). Quality of life 693 was not described and no significant harms were identified. There are observational 694 cohort studies of topical glucocorticosteroids and other maintenance treatment options 695 which provide some additional evidence. 696 697 We found 6 single-armed observational cohorts for topical glucocorticosteroids. Butz et 698 al 2014(20) reported that 11 out of 15 were able to maintain remission on a lower dose 699 of topical glucocorticosteroid (fluticasone 0.88 mg) over 3 months. Andreae et al 2016 700 (86) reported on 54 pediatric patients treated long term with swallowed fluticasone and 701 with mean follow-up of 20 months found 63% remained in histologic remission. Dellon 702 et al 2016 (87) reported in an abstract that 42% of their cohort were able to maintain 703 remission on budesonide 2 mg/day. Greuter et al 2017(88) reported that out of 33 704 people who had achieved clinical and histological remission for 6 months, 27 705 experienced relapse with an average time-to-relapse of 22 weeks after stopping topical 706 glucocortiocsteroids. Eluri et al 2017 (89) reported that 20 out of 33 adults who were 707 using topical glucocorticosteroids experienced a relapse when followed over a 12 month 708 period. Rubenstein et al 2018 ***reported that 7 of 8 children who attempted to reduce 709 budesonide to a 3-times-per-week dosing schedule from a daily schedule experienced a 710 relapse. 711 712 Alexander et al 2017 (90) was profiled earlier in this technical report, and is listed here 713 because the subjects were in remission when they were randomized to montelukast or 714 placebo. 715 716 We identified 3 long term PPI studies that reported remission/relapse rates over 717 extended time periods. Molina-Infante 2015(91) reported that 55 of 75 remained in 718 remission on PPI with a mean follow-up length of 26 months. Gomez-Torrijos 2016(40) 719 reported that 31 of 38 remained in remission when dose of PPI reduced to once daily, 720 and 15 of 18 remained in remission when daily high dose PPI was reduced to regular 24

721 dose PPI. Gutierrez-Junquera 2016(49) reported that 17 of 57 failed to maintain 722 remission over a 1 year period on 1 mg/kg per dose twice daily of PPI. 723 724 We identified 3 single-armed cohorts of long term dietary treatment. Lucendo 2013 725 (92) reported that 25 of 42 who had initially achieved remission with dietary therapy 726 remained in remission 52 weeks later, with many patients dropping out of the study. 727 Philpott 2015 (93) reported that 10 of 10 who maintained dietary therapy remained in 728 remission with a mean follow-up length of 36 weeks. Reed 2017 (72) reported that 10 729 of 10 who maintained dietary therapy remained in remission over a mean follow-up 730 length of 25 weeks. 731 732 Overall, it appears clear from the placebo arms of randomized trials, natural history 733 studies, and cohort studies that if treatments in EoE are stopped, then disease activity 734 (including symptomatic, endoscopic, and histologic) has a high chance of recurring. 735 The difficulty is that there are few data to guide either treatment or surveillance of long 736 term treatment in EoE. A general approach is to repeat an endoscopy for monitoring 737 after treatment changes, which is discussed in a later question in this document. For 738 dietary treatment, continuing to avoid confirmed food triggers should be effective but 739 may have significant nutritional and/or quality of life deficits depending on the nature 740 and duration of prolonged avoidance. However, the details of dosing, treatment 741 intensity, and endoscopic surveillance frequency remain areas that need to be studied. 742 Maintenance Therapy compared to placebo for eosinophilic esophagitis

Bibliography: Straumann 2011a (50)

Outcomes and Follow № of Certainty Relative Anticipated absolute up participants of the effect effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with Maintenance Therapy

Not achieving histologic 28 ⨁◯◯◯ RR 0.70 714 per 214 fewer remission (Eos < 20/hpf) (1 RCT) VERY (0.38 to 1,000 per 1,000 (Hist remission) LOWb,c 1.30) (443 fewer to assessed with: Eos < 214 more) 20/hpf follow up: mean 50 weeks 25

Maintenance Therapy compared to placebo for eosinophilic esophagitis Bibliography: Straumann 2011a (50)

Outcomes and Follow № of Certainty Relative Anticipated absolute up participants of the effect effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with Maintenance Therapy

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

743 Explanations 744 745 a. To be eligible had to be in remission eos < 5 /hpf on high dose budesonide 2 mg/day 746 and low symptom score (< 2 on their scale); budesonide given via TIA nebulizer (does 747 not exist in US) at dose 0.5 mg (low); partial remission < 20 rather than < 15 which is 748 used for most other intervention outcomes; not all patients completed 50 weeks---if 749 clinical symptoms and relapse confirmed they stopped (9 completed 50 weeks in BUD 750 group and 5 in PLAC) 751 b. Used low dose of budesonide delivered through a device not available in the US 752 c. Very low information size and CI crosses 1. 753 754 #8 Should esophageal dilation be used in patients with EoE? 755 756 Evidence Summary: Histologic remission was not assessed for this intervention. 757 Estimates were taken from the Dougherty 2017 meta-analysis(94) that investigated the 758 use of dilation in patients with EoE. Data for outcomes of interest were extracted and 759 pooled from studies included in the Dougherty meta-analysis that explicitly noted that 760 they were performed with greater than 5 participants. We summarized outcomes on 761 clinical improvement as well as adverse events (mortality, perforation, hospitalization, 762 and hemorrhage). Rates for each outcome are presented. The assumption was that no 763 clinical improvement nor adverse events would reasonably occur if dilation was not 764 performed. 87% of patients experienced clinical improvement with esophageal dilation 765 in symptoms (but not esophageal eosinophil counts). There was no mortality associated 766 with dilation. The pooled perforation rate was 0.4%, hospitalization was reported after 767 1.2% of dilations, and significant GI hemorrhage was reported in 0.1% of dilations. 26

768 769 Quality of Evidence: The certainty in the effect estimate was very low across all 770 outcomes. We rated down for risk of bias given that there was no control group. This, 771 combined with the fact that we started with a majority of observational data, yielded very 772 low certainty in the effect. It is also important to note that the assessment of clinical 773 improvement does not account for concomitant use of medication or diet. We did not 774 rate down for indirectness though it was noted that patients who need dilation have 775 fibro-stenotic disease. Though this population may be distinct from those included in 776 studies where therapeutic management with medications was investigated, for “clinical 777 improvement”, we did not rate down for inconsistency despite heterogeneity of the 778 pooled estimate (I2); our assumption was that dilation does indeed result symptomatic 779 improvement. 780 781 Discussion: There are 3 meta-analyses from 2017 that investigated the use of dilation in 782 patients with EoE (94-96). We compared them, found the Dougherty to be the most 783 inclusive after discussion with the expert panel, and used that study as the basis for our 784 evidence profile. Use of dilation in this patient population was not associated with any 785 noted safety risks. Clinicians should recognize that dilation is not a treatment for the 786 inflammation associated with EoE per se, but rather a treatment directed at the 787 dysphagia symptoms associated with EoE. Histologic outcomes are not routinely 788 reported in dilation studies nor are biopsies taken during dilation, and this measure is 789 not being discussed in the context of a management strategy that would decrease 790 esophageal eosinophilia. We recognized that there was significant variability in how 791 several outcomes were measured in the constituent studies. Rates and absolute effects 792 were presented because the majority of the included studies had no control group. 793 794 27

Esophageal dilation compared to no dilatation for Eosinophilic Esophagitis Bibliography: Dougherty 2017(94)

Outcomes № of Certainty Relative Anticipated absolute participants of the effect effects (studies) evidence (95% CI) Risk with Risk Follow-up (GRADE) no difference dilatation with Esophageal dilation

Clinical Improvement 1928 ⨁◯◯◯ not 0 per 100 0 fewer per (not defined) a,b dilations VERY estimable g 100 (14 LOW d,e,f (0 fewer to 0 observational fewer) studies) c

Mortality h 2772 ⨁◯◯◯ not -- per -- g -- per -- dilations VERY estimable (-- to --) (20 LOW d,f observational studies) i

Perforation h 2772 ⨁◯◯◯ not 0 per 0 fewer per dilations VERY estimable 1,000 g 1,000 (20 LOW d,f (0 fewer to 0 observational fewer) studies) i

Hospitalization h 2466 ⨁◯◯◯ not 0 per 0 fewer per dilations VERY estimable 1,000 g 1,000 (12 LOW d,f (0 fewer to 0 observational fewer) studies) j 28

Esophageal dilation compared to no dilatation for Eosinophilic Esophagitis Bibliography: Dougherty 2017(94)

Outcomes № of Certainty Relative Anticipated absolute participants of the effect effects (studies) evidence (95% CI) Risk with Risk Follow-up (GRADE) no difference dilatation with Esophageal dilation

Hemorrhage h,k 2588 ⨁◯◯◯ not 0 per 0 fewer per dilations VERY estimable 1,000 g 1,000 (12 LOW d,f (0 fewer to 0 observational fewer) studies)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

795 Explanations 796 a. Follow up Median of 12 months and a range from 1 week to 36 months 797 b. Clinical improvement expressed per patient. Heterogeneous definition of "clinical 798 improvement". 799 c. 37 studies were included in meta-analysis, though only 14 studies were included for 800 this outcome. There was one randomized trial (Kavitt) and there rest were observational 801 studies with greater than 5 participants. Some studies did not specify the number of 802 participants, and as such, were excluded. The included case-control (Cohen et al) study 803 in the Dougherty meta-analysis was also excluded in this analysis. 804 d. Absolute rate calculated from data provided in Dougherty et al meta-analysis, pooling 805 data from cohort/RCT studies where "n" was available for outcome of interest. Case- 806 control and observational studies will <5 patients were excluded. Only absolute rates 807 provided because there was no control group. 808 e. Despite heterogeneity of pooled estimate, assumption is that dilation does cause 809 symptomatic improvement. 29

810 f. It was noted that patients who need dilation have fibrostenotic disease. Though this 811 population may be distinct from those included in studies where therapeutic 812 management with medications was investigated, we did not rate down for indirectness. 813 g. Assumption was that no clinical improvement nor adverse events could occur if 814 dilation not performed. 815 h. Expressed as number per dilation 816 i. 37 studies were included in meta-analysis, though only 20 studies were included for 817 this outcome. There was one randomized trial (Kavitt) and there rest were observational 818 studies with greater than 5 participants. Some studies did not specify the number of 819 participants, and as such, were excluded. The included case-control (Cohen et al) study 820 in the Dougherty meta-analysis was also excluded in this analysis. 821 j. 37 studies were included in meta-analysis, though only 12 studies were included for 822 this outcome. There was one randomized trial (Kavitt) and there rest were observational 823 studies with greater than 5 participants. Some studies did not specify the number of 824 participants, and as such, were excluded. The included case-control (Cohen et al) study 825 in the Dougherty meta analysis was also excluded in this analysis. 826 k. Significant GI bleeds were those defined as needing additional clinical intervention, 827 usually re-endoscopy, and did not include mucosal tears seen immediately after dilation. 828 829 #9 Should anti-IL-5 therapy be used in patients with EoE? 830 831 Evidence Summary: There are 3 RCTs of anti-IL-5 therapy compared to placebo. Anti- 832 IL-5 treatment had little or no effect, with 94.4% of patients assigned to anti-IL-5 833 therapy failed to achieve histologic remission compared to 93.9% of the placebo group, 834 for a RR of 0.92, 0.84, 1.00. 835 836 Quality of Evidence: The certainty in the effect estimate was low. The certainty in the 837 estimate was downgraded for indirectness because the participants in this study were 838 different than other interventions in that many had failed the other interventions prior to 839 entering the trials. The certainty was also downgraded for imprecision because the 840 confidence interval included 1. 841 842 Discussion: Very few individuals in the intervention or placebo arms achieved the pre- 843 specified histologic remission rate of < 15 eosinophils/hpf. We grouped 2 different 844 drugs with similar mechanisms of action-mepolizumab and reslizumab-for the effect size 845 estimate. One of the studies (Assa’ad 2011) did not have a true placebo group but 846 instead used a low dose of mepolizumab as a comparator. The participants in these 847 studies frequently failed other treatments and had higher levels of esophageal 848 eosinophilia upon entry into the study than in studies for other interventions. Symptom 849 outcomes were reported differently in the 3 studies and therefore were not grouped to 850 create an effect estimate. Quality of life was reported in 1 study (Spergel 2012), and 851 there were no major signals of harms reported in the 3 studies. 852 853 . 30

Anti-IL-5 monoclonal antibodies compared to placebo for eosinophilic esophagitis Bibliography: Straumann 2010a,(97) Assa'ad 2011b,(98) Spergel 2012c (99)

Outcomes and Follow- № of Certainty Relative Anticipated absolute up participants of the effect effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with Anti- IL-5 monoclonal antibodies

Not achieving histologic 286 ⨁⨁◯◯ RR 0.92 902 per 72 fewer remission ( < 15 eos/hpf) (3 RCTs) LOWe,f (0.84 to 1,000 per 1,000 (Remission (partial)) 1.00) (144 fewer assessed with: < 15/hpf to 0 fewer) follow up: range 9 weeks to 16 weeks

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

854 Explanations 855 a. Mepolizumab in adults (escalating mepolizumab doses) 856 b. Mepolizumab in children, used very low dose mepolizumab as placebo or 857 “comparator” group as suggested by authors in the methods section; grouped other 2 858 doses of mepolizumab together (no obvious dose response); missing description of 859 blinding and allocation concealment 860 c. Reslizumab in children; grouped 3 doses of reslizumab together in intervention 861 group (no obvious dose response) 862 d. Participant selection for these studies are different than for other interventions— 863 have failed many other interventions prior to enrolling in this study 864 e. CI crosses 1 865 f. CI touches 1 (upper boundary of CI is 1.0—no effect—which could lead to a 866 different recommendation and therefore represents imprecision. 31

867 868 #10 Should anti-IL-13 therapy be used in patients with EoE? 869 870 Evidence Summary: There is 1 published RCT of anti-IL-13 therapy compared to 871 placebo. The RR was 1.08 (95% CI 0.81-1.40) for participants in the anti-IL-13 arm 872 failing to achieve histologic remission compared to placebo. 873 874 Quality of Evidence: The certainty in the effect estimate was low. The certainty in the 875 estimate was downgraded for imprecision due to very low information size and for 876 indirectness as patients who entered the study were more likely to have failed other 877 treatments and have very high baseline esophageal eosinophilia. 878 879 Discussion: A single small RCT failed to show or exclude an effect of anti-IL-13 in EoE. 880 Quality of life was not described and no significant harms were reported. The expert 881 panel identified 2 additional RCTs which are reported as abstracts but not yet published 882 as full manuscripts. These 2 studies each included interventions that are somewhat 883 different than in Rothenberg et al. In Hirano et al, dupilumab, an IL-13/IL-4 receptor 884 blocker, was compared with placebo. In the dupilumab arm, 4 of 23 failed to achieve 885 histologic remission < 15/hpf compared to 24 out of 24 in the placebo arm (risk 886 ratio=0.17, 0.07-0.42). In Dellon et al, two doses of an anti-IL-13Ra1/Ra2 blocker, was 887 compared to placebo in 99 subjects. The mean eosinophil counts were significant 888 reduced from baseline for both doses levels compared to placebo (-99.9 eos/hpf for 889 high dose, -94.8 eos/hpf for low dose, and -4.4 eos/hpf for placebo; all comparisons 890 p<0.0001). 891 892 a. QAX-576 compared to placebo for eosinophilic esophagitis

Bibliography: Rothenberg 2015 (100)

Outcomes and Follow-up № of Certainty Relative Anticipated participants of the effect absolute effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with QAX- 576

Not achieving histologic 25 ⨁⨁◯◯ RR 0.94 875 per 53 fewer remission < 15 eos/hpf (hist (1 RCT) LOWa,b (0.67 to 1,000 per 1,000 remission) 1.30) (289 fewer assessed with: < 15 eos/hpf to 263 follow up: mean 12 weeks more) 32

QAX-576 compared to placebo for eosinophilic esophagitis Bibliography: Rothenberg 2015 (100)

Outcomes and Follow-up № of Certainty Relative Anticipated participants of the effect absolute effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with QAX- 576

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

893 Explanation 894 a. Very low information size 895 b. Patients with higher baseline esophageal eosinophilia and more likely to have 896 failed other treatments 897 898 #11 Should anti-IgE therapy be used in patients with EoE? 899 900 Evidence Summary: There is 1 RCT of anti-IgE therapy compared to placebo. The RR 901 was not estimable because no subjects in either trial arm achieved histologic remission. 902 903 Quality of Evidence: The certainty in the effect estimate was very low. The certainty in 904 the estimate was downgraded for imprecision due to very low information size and for 905 indirectness because subjects were selected who failed topical steroid. 906 907 Discussion: A single very small RCT showed no effect of omalizumab in EoE. We 908 identified an observational cohort (Loizou 2015) (101) but elected to only consider the 909 RCT given the stronger study design and larger overall number of subjects. Quality of 910 life or harms were not described in the RCT. 911 912 33

Omalizumab compared to placebo for eosinophilic esophagitis Bibliography: Clayton 2014 (102)

Outcomes and № of Certainty Relative Anticipated absolute Follow-up participants of the effect effects (studies) evidence (95% Risk with Risk (GRADE) CI) placebo difference with Omalizumab

Change in peak 30 ⨁◯◯◯ - The mean MD 6 eosinophil counts/hpf (1 RCT) VERY change in eos/hpf (Peak eosinophils) LOWa,b peak more assessed with: eos/hpf eosinophil (4 fewer to follow up: mean 16 counts/hpf 16 more) weeks was 0 eos/hpfc

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; MD: Mean difference

913 a. Selected subjects who failed topical steroid; some had trialed PPI before and 914 some were on PPI during trial, outcome is change in mean eosinophils which may be a 915 less direct outcome to consider. 916 b. Very low information size 917 c. Personal communication with Dr. Fred Clayton that no subject in either arm 918 achieved histologic remission < 15 eos/hpf. 919 920 #12 Should montelukast be used in patients with EoE? 921 922 Evidence Summary: There is 1 RCT that compares montelukast with placebo for 923 maintenance therapy after subjects had achieved symptomatic and histologic 924 remission(90), and 4 observational studies that report outcomes after montelukast use 925 [Atwood 2003 (n=8)(103); Vanderhoof 2003 (n=8)(104); Lucendo 2011 (n=11)(105); 926 Stumphy 2011 (n=8)(106)]. Based on the RCT, the RR for the recurrence of solid food 927 dysphagia was 0.79 (95% CI 0.51 to 1.21). Failing to achieve histologic remission of < 928 15 eosinophils/high power field was not measured in this trial. 929 34

930 Quality of Evidence: The certainty in the effect estimate was very low. The certainty 931 was downgraded for serious indirectness because subjects in the study had already 932 achieved remission with topical glucocorticosteroid and for very serious imprecision due 933 to very low information size. The observational data were not summarized in an 934 evidence profile. 935 936 Discussion: The findings from the RCT are most relevant to patients who had already 937 achieved remission after taking topical glucocorticosteroids. However, the certainty 938 about the efficacy of montelukast for EoE after patients achieved remission with topical 939 glucocorticosteroid is very low. The findings are not informative for the outcome of 940 histological remission because this outcome was not measured in the single clinical trial 941 of montelukast for EoE. The trial was performed in adults and therefore the data may 942 not be applicable to children. The findings are not directly relevant for patients who are 943 newly diagnosed with EoE and have not started any previous treatments. 944 945 Montelukast compared to placebo for management of Eosinophilic esophagitis

Bibliography: Alexander 2017 (90)

Outcomes and Follow- № of Certainty Relative Anticipated absolute up participants of the effect effects (studies) evidence (95% Risk with Risk (GRADE) CI) placebo difference with montelukast

Solid food dysphagia 41 ⨁◯◯◯ RR 0.79 760 per 160 fewer (improvement) (1 RCT) VERY (0.51 to 1000 per 1000 (Telephone symptom LOW b,c, 1.21) (370 fewer to questionnairre)a 160 more) follow up: 26 weeks

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio 35

946 Explanations 947 a. The abridged dysphagia questionnaire used the questions on dysphagia frequency, 948 severity, and food impaction from the Mayo Dysphagia Questionnaire, 2-week version 949 b. All patients had first achieved remission with topical glucocorticosteroid treatment. 950 Therefore, this is indirect to our PICO question of should we use montelukast for EoE as 951 the question is very specific to should we use montelukast to maintain remission. 952 c. Large confidence interval that crosses 1; estimate based on a single RCT 953 954 955 #13 Should cromolyn be used in patients with EoE? 956 957 Evidence Summary: There is 1 RCT of cromolyn compared to placebo. Based on the 958 trial, 89% of subjects (8 out of 9) treated with cromolyn failed to achieve histologic 959 remission compared to 100% (7 out of 7) in the placebo arm. 960 961 Quality of Evidence: The certainty in the effect estimate was low. The certainty in the 962 estimate was downgraded twice for very serious imprecision given the very low number 963 of study participants and because the 95% confidence interval crosses 1. 964 965 Discussion: Mast cells are implicated in EoE pathogenesis. Therefore, targeting mast 966 cells with cromolyn has biological plausibility. The single, small study performed with 967 cromolyn does not exclude the possibility of a benefit for cromolyn in patients with EoE. 968 An observational study of 14 children with EoE treated with cromolyn found that none of 969 the children had improvement in histology or symptoms. (59) 970 971 Cromolyn compared to placebo for eosinophilic esophagitis Bibliography: Lieberman 2018(107)

Outcomes and Follow-up № of Certainty Relative Anticipated participants of the effect absolute effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with cromolyn

Not achieving histologic 16 ⨁⨁◯◯ RR 0.89 1,000 per 110 fewer remission (Histologic (1 RCT) LOWa (0.71 to 1,000 per 1,000 remission) 1.12) (290 fewer assessed with: Eos < 5/hpf to 120 follow up: mean 8 weeks more) 36

Cromolyn compared to placebo for eosinophilic esophagitis Bibliography: Lieberman 2018(107)

Outcomes and Follow-up № of Certainty Relative Anticipated participants of the effect absolute effects (studies) evidence (95% Risk Risk (GRADE) CI) with difference placebo with cromolyn

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

972 Explanation 973 a. Very serious imprecision 974 975 #14 Should anti-TNFs be used in pts with EoE? 976 977 Evidence Summary: There is one observational study, a case-series described as an 978 open-label, nonrandomized pilot T1 translational trial, that investigated the use of 979 infliximab as acute therapy in 3 adults with EoE who were steroid-resistant and included 980 patients had active EoE.(108) Three outcomes were measured: Response (as inferred 981 by study report of esophageal eosinophilic infiltration), Response (as inferred by 982 symptom score (Straumann's Criteria), and endoscopic alterations (Straumann's 983 Criteria). Results are described narratively because quantitative summary estimates 984 were not presented in the included study. 985 986 Quality of Evidence: The overall certainty in the effects was very low. The certainty was 987 rated down due to risk of bias (no control population, the possibility of selection bias. 988 and that fact outcome measures may not be well-validated). We could not assess 989 publication bias, effect size, or confounding. It is a relevant patient population, though 990 one that is refractory to standard therapy. We did rate down for indirectness given that it 991 was a population that was refractory to standard therapy. 992 993 Discussion: Active EoE was defined clinically as dysphagia (when not on anti- 994 inflammatory therapy) and histologically by a peak cell density of greater than 24 995 eosinophils/ high-power field (is this also on anti-inflammatory therapy?). Included 996 patients had “inadequate response to prior treatment”. This was explained by the 997 authors as Patients 1 and 3 were almost free of symptoms during maintenance therapy 37

998 with topical fluticasone, but immediately after cessation of the medication, symptoms 999 reappeared. Patient 2 had been receiving systemic corticosteroid treatment for the past 1000 8 years and needed at least 10 mg of prednisone per day for symptom control. Patients 1001 2 and 3 had previously undergone repeated dilations for the treatment of strictures. No 1002 adverse events were reported. 1003 1004 1005 #15 Should immunomodulators be used in the treatment of EoE? 1006 1007 Evidence Summary: Two observational studies were included identified that 1008 investigated the use of immunomodulators in EoE. The outcome listed in the evidence 1009 profile was response though it was variably defined in the included studies as 1010 “symptomatic remission” or “clinical remission”. Results are described narratively 1011 because quantitative summary estimates were not presented in the included study. 1012 1013 Quality of the Evidence: The certainty in the effect estimate was very low. The quality of 1014 the evidence was rated down for the lack of control populations, suspected selection 1015 bias, possible confounding, and outcomes that were not well-defined. There was also 1016 concern regarding indirectness of the included patient population, as a more severe 1017 disease phenotype, given that included patients were steroid-dependent. We did rate 1018 down for indirectness given that it was a population that was refractory to standard 1019 therapy. 1020 1021 1022 1023 Discussion: The evidence was derived from 2 papers with a total included population of 1024 4 people; therefore the evidence base for immunodulator treatment in patients with EoE 1025 is very small.(109, 110) In these case series, it is not clear if the histologic changes 1026 were related to starting the immunomodulator or other factors, such as the attempts with 1027 withdraw systemic steroid. Therefore it is difficult to draw any conclusions about how 1028 immunomodulators work for EoE. 1029 1030 1031 Narrative Summaries 1032 1033 The following 2 PICO questions are addressed as narrative summaries based on the 1034 consensus of the expert panel that data in format amenable to GRADE analysis was not 1035 available. 1036 #16 Should repeat EGD be used to assess patients with EoE after a change in 1037 treatment? 1038 1039 The role of performing upper endoscopy and biopsy for monitoring EoE biologic disease 1040 activity (endoscopic severity assessed visually with the EREFS classification and 1041 esophageal eosinophilia assessed histologically) has not been formally studied in higher 1042 quality trials. However, there are numerous studies that support performing endoscopy 38

1043 to survey disease activity. This is based on several concepts. The first is an 1044 understanding that for many patients, the natural history of untreated esophageal 1045 eosinophilia is a progression from an inflammatory to a fibrostenotic phenotype(1). 1046 Cohort studies of untreated patients and placebo groups of RCTs repeatedly 1047 demonstrate that esophageal eosinophilia does not resolve over time and that patients 1048 with EoE do not “grow out of it”(111). Further, a set of four studies from different 1049 centers in the U.S. and Europe independently show that the longer the duration of 1050 disease prior to diagnosis (as a proxy for time prior to treatment), the higher the 1051 proportion of patients who have a stricture or fibrostenotic phenotype at the time of 1052 diagnosis (12-15). For example, in one study, more than 80% of patients had strictures 1053 if the diagnostic delay was >20 years. Other studies document progression in distinct 1054 patients (112). Therefore, because there is a consequence to persistent eosinophilic 1055 inflammation in the esophagus (fibrosis leading to strictures), it is important to survey to 1056 confirm that esophageal eosinophilia has been corrected. Second, symptoms only 1057 modestly correlate with disease activity. Numerous studies have shown discordance 1058 between symptoms of esophageal dysfunction and endoscopic and histologic disease 1059 activity (18, 75, 113). There are several reasons for this. Patients can avoid foods that 1060 cause dysphagia or other symptoms, or modify the way they eat (chewing carefully, 1061 eating slowly, lubricating foods, drinking copious fluids) to minimize symptom regardless 1062 of the level of biologic disease activity. Additionally, symptoms and biologic activity do 1063 not have a linear relationship, and symptoms may remain quite mild until a certain 1064 threshold of endoscopic severity is reached. If patients have previously had an 1065 esophageal stricture and have undergone esophageal dilation, then symptoms of 1066 dysphagia will be improved regardless of underlying biologic activity. The third concept 1067 is specific for PPIs. Reflux and EoE can have a complicated relationship. In particular, 1068 the two conditions may overlap, and EoE may predispose to secondary reflux or less 1069 effective clearance of physiologic reflux (due to decreased esophageal compliance 1070 and/or the mild dysmotility)(114). In this situation, PPIs may help to treat reflux and thus 1071 improve some symptoms, but might not be effective for the underlying EoE. In sum, the 1072 chronic nature of EoE where esophageal eosinophilia can lead to progressive fibrosis, 1073 the potential discordance between symptoms and underlying biologic disease activity of 1074 EoE, and a possible non-EoE-related mechanism of potential symptom response to 1075 PPI, all suggests that endoscopy for surveillance may be effective, even in patients who 1076 have a symptom response to PPI. Reasons for not performing follow up endoscopies 1077 include potential risks of sedation impacting development in younger children, risks of 1078 repeated endoscopy, financial and time burden. These considerations need to be 1079 accounted for when balancing the risks and benefits of performing surveillance 1080 endoscopies especially in children. 1081 1082 1083 1084 #17 What is the management of patients who become asymptomatic after initial 1085 PPI treatment? 1086 1087 39

1088 The significance of esophageal eosinophilia is not a pathognomonic finding and has to 1089 be carefully considered within the appropriate clinical context. For instance, if an 1090 otherwise healthy atopic adult patient undergoes endoscopy for food impaction or 1091 dysphagia and is found to have esophageal eosinophilia, both the literature and clinical 1092 experience support a probable diagnosis of eosinophilic esophagitis and high likelihood 1093 of some response to PPIs (See Question 1). If a young child with chronic vomiting, 1094 abdominal pain and weight loss is found to have esophageal eosinophilia, a diagnosis 1095 of gastroesophageal reflux is more probable and similarly will have a high likelihood of 1096 response to PPI. Endoscopic features and associated histological findings beyond 1097 eosinophil counts should be considered supplementary to helping establish diagnostic 1098 clarity. A friable mucosa is an unusual finding in EoE, whereas extensive eosinophilic 1099 degranulation would be a much more common finding. 1100 1101 In either situation, an argument can be made for the value of a post PPI treatment 1102 follow-up endoscopy. For either diagnostic situation detailed above, resolution of 1103 eosinophilia would need to be documented to ensure healing has occurred that may 1104 potentially alter the long-term outcome. Historically, children with EoE have 1105 demonstrated a poor correlation between symptoms and inflammation. Recent data 1106 from Aceves et al has recently showed that proximal eosinophilia associates with self- 1107 reported symptoms in a sample of children from CeGIR consortium centers, which may 1108 hold potential promise as a marker to monitor disease progression.(115) 1109 1110 The long term management of children and adults with esophageal eosinophilia who are 1111 treated with PPIs remains an evolving concept. If a patient is thought to have EoE, long 1112 term treatment would be indicated with appropriate follow up as described above and in 1113 Question #16. Clinical observation for side effects of PPIs is warranted. While a degree 1114 of overlap exists between EoE and GERD (116, 117), performance of a fundoplication 1115 for PPI responsive EoE is not currently recommended because of the potential anti- 1116 inflammatory properties of PPIs, the theoretical concern for retention of offending food 1117 antigens in the esophagus, and worsening dysmotility due to a tight gastro-esophageal 1118 junction. 1119 1120 There are multiple unresolved additional issues, including establishing the optimal 1121 minimal duration of PPI treatment prior to repeat endoscopy, optimal dose and duration 1122 of PPI use as a primary EoE treatment, optimal duration of long-term PPI treatment if a 1123 PPI response is observed, and determining the next best treatment if inflammation 1124 persists despite PPI therapy. 1125 1126 1127 1128 SUMMARY AND CONCLUSIONS 1129 1130 In evaluating the efficacy of multiple treatments for EoE to achieve a primary outcome of 1131 reducing esophageal eosinophil counts to <15 eos/hpf, few treatments have moderate 1132 certainty of evidence for such an effect, and most have low to very low certainty of 1133 evidence for such an effect. There is moderate certainty in the evidence that topical 40

1134 glucocorticosteroids effectively reduce esophageal eosinophil counts to <15/hpf over a 1135 short-term treatment period of 4-12 weeks, but very low certainty about the effects of 1136 using topical glucocortiocsteroids as maintenance therapy given the lack of studies on 1137 this topic. Moderately certain evidence suggests that systemic glucocorticosteroids 1138 have similar efficacy rates as topical glucocorticosteroids, but at a cost of higher rates of 1139 adverse effects. Multiple dietary strategies may be effective in reducing esophageal 1140 eosinophil counts to <15/hpf over a short-term treatment period, with moderate certainty 1141 for elemental diets, low certainty for empiric 4 and 6 food elimination diets, and very low 1142 certainty for allergy-based testing dietary eliminations and empiric 1 and 2 food 1143 elimination diets. We report very low certainty for the effect of PPIs in patients with 1144 esophageal eosinophilia and for the effects of esophageal dilation in patients with EoE, 1145 although it appears to be relatively safe. We found low or very low certainty in the 1146 effects of multiple other medical treatments for EoE: anti-IL-5 therapy, anti-IL-13 1147 therapy, anti-IgE therapy montelukast, cromolyn, and anti-TNF therapy, many of which 1148 failed to exclude or confirm a benefit. Current research should focus on directly 1149 comparing available treatments with more reliable study designs, testing new 1150 treatments, using validated symptom questionnaires, and consistently measuring quality 1151 of life and nutritional status. 1152 1153 APPENDIX 1154 1155 Table of Contents: 1156 Table 1. PICO Questions 1157 Table 2. Summary of EoE outcomes considered 1158 Table 3. Electronic Search Strategies 1159 Table 4. Additional Evidence Profiles 1160 Figure 1. PRISMA Flow diagram for identifying articles 1161 Figure 2. Forest Plots for each PICO Question 1162 1163 1164 Appendix Table 1. PICO Questions 1165 # Question PICO Question Method Population Intervent Comparato Outcome ion(s) r 1 Should PPIs Esophageal PPI Placebo -Failing to GRADE be used in eosinophilia, achieve patients with not histologic symptomatic necessarily remission esophageal meeting EoE (eosinophils<15 eosinophilia? criteria /hpf) -Symptoms -Adverse events

41

2 Should topical Patients with Topical Placebo -Failing to GRADE glucocortiocste EoE Steroids achieve roids be used histologic in patients with remission EoE? (eosinophils<15 /hpf) -Symptoms -Adverse events

3 Should Patients with Systemic Placebo -Failing to GRADE systemic EoE steroids achieve glucocorticoste or histologic roids be used remission in patients with topical (eosinophils<15 EoE? steroids /hpf) -Symptoms -Adverse events

4 Should an Patients with Elementa No diet -Failing to GRADE elemental diet EoE l diet change achieve be used in histologic patients with remission EoE? (eosinophils<15 /hpf) -Symptoms -Adverse events

5 Should an Patients with -6 food No -Failing to GRADE empiric EoE eliminatio elimination achieve elimination diet n diet diet histologic be used in -4 food remission patients with eliminatio (eosinophils<15 EoE? n diet /hpf) -2 food -Symptoms eliminatio -Adverse n diet events -Single food eliminatio n diet

42

6 Should allergy- Patients with -SPT, No -Failing to GRADE based testing EoE sIGE, elimination achieve be used for the APT, diet histologic purpose of and/or remission identifying CRD to (eosinophils<15 food triggers in direct an /hpf) patients with eliminatio -Symptoms EoE? n diet -Adverse events

7 Is Patients with Maintaini Discontinui -Failing to GRADE maintenance EoE who are ng ng therapy achieve therapy in clinical therapy histologic recommended and remission in patients with histologic (eosinophils<15 EoE? remission /hpf) -Symptoms -Adverse events

8 Should Patients with Dilation No dilation -Adverse GRADE esophageal EoE events dilation be used in patients with EoE? 9 Should anti-IL- Patients with Anti-IL-5 Placebo -Failing to GRADE 5 therapy be EoE therapy achieve used in histologic patients with remission EoE? (eosinophils<15 /hpf) -Symptoms -Adverse events 1 Should anti-IL- Patients with Anti-IL-13 Placebo -Failing to GRADE 0 13 therapy be EoE therapy achieve used in histologic patients with remission EoE? (eosinophils<15 /hpf) -Symptoms -Adverse events 43

1 Should anti- Patients with Anti-IgE Placebo -Failing to GRADE 1 IgE therapy be EoE therapy achieve used in histologic patients with remission EoE? (eosinophils<15 /hpf) -Symptoms -Adverse events 1 Should Patients with Monteluk Placebo -Failing to GRADE 2 montelukast EoE ast achieve be used in histologic patients with remission EoE? (eosinophils<15 /hpf) -Symptoms -Adverse events 1 Should Patients with Cromolyn Placebo -Failing to GRADE 3 cromolyn be EoE achieve used in histologic patients with remission EoE? (eosinophils<15 /hpf) -Symptoms -Adverse events 1 Should anti- Patients with Anti-TNF Placebo -Failing to GRADE 4 TNFs be used EoE achieve in patients with histologic EoE? remission (eosinophils<15 /hpf) -Symptoms -Adverse events 1 Should Patients with Azathiopr Placebo -Failing to GRADE 5 Immunomodul EoE ine or 6- achieve ators be used MP histologic in patients with remission EoE? (eosinophils<15 /hpf) -Symptoms -Adverse events 44

1 Should repeat Patients with EGD with N/A N/A Narrative 6 EGD be used EoE Esophag to assess eal patients with biopsy EoE after a change in treatment?

1 Management Patients with Repeat No EGD, -Failing to Narrative 7 of patients esophageal EGD, no change achieve who become eosinophilia followed in treatment histologic asymptomatic and/or EoE by remission on initial PPI additional (eosinophils<15 treatment therapy if /hpf) eosinophi -Symptoms ls still -Adverse elevated events 1166 1167 1168 1169 1170 Appendix Table 2. Summary of EoE outcomes considered 1171 Name of Outcome Definition Minimally Rating Critical Important (1-9) vs Difference Important vs Non- important Histology eosinophil Not known 8 Critical counts <15/hpf in esophagus Symptoms DSQ, MDQ, Response: 30-50% 6 Important EEsAI, PEES improvement Remission: Absence of symptoms. For EEsAI, score <30 may indicate inactive disease/remission Endoscopic Findings (EREFS) Response: 30-50% 6 Important improvement 45

Remission: normalization Complications of EoE Stricture N/A 8 Critical Food impaction Malnutrition psychiatric Quality of Life Peds EoE QL Response: 30-50% 6 Important Adult Eoe QL improvement as response Remission: Baseline for healthy people Biomarkers Peripheral Eos N/A 4 Non- Progenitor Eos important Gene Expression Cytosponge EST readout Tissue cytokines eosinophil granules Adverse Events of Allergy N/A 6 Important Medication Treatment Weight gain Anaphylaxis Infections Complications of Perforation N/A 8 Critical endoscopy Anesthesia complications Costs N/A 6 Important 1172 1173 Appendix Table 3. Electronic Search Strategies 1174 1175 Search Strategy 1176 EoE – Final Search 1177 1178 Search date: May 13, 2018 1179 Databases searched: Ovid MEDLINE: Epub Ahead of Print, In-Process & Other Non- 1180 Indexed Citations, Ovid MEDLINE® Daily and Ovid MEDLINE® 1946-Present, Embase 1181 Classic+Embase 1947 to 2018 May 11; Wiley Cochrane. Individual systematic review 46

1182 searches were not registered in Prospero, as this technical report represents an 1183 umbrella evidence review. 1184 1185 Ovid MEDLINE(R), Embase # Searches Result s 1 exp Eosinophilic Esophagitis/ 5991 2 (eosinophilic adj2 (esophagitis or oesophagitis)).ti,ab. 6556 3 1 or 2 7480 4 exp EOSINOPHILIA/ 62171 5 exp ESOPHAGITIS/ 45445 6 exp ESOPHAGUS/ 133201 7 4 and (5 or 6) 2855 8 eoe.ti,ab. 3588 9 3 and (7 or 8) 4580 10 exp Adrenal Cortex Hormones/ use ppez 376133 11 exp corticosteroid/ use emczd 935731 12 (corticosteroid* or steroid* or prednisone or budesonide or fluticasone).ti,ab.786701 13 exp Leukotriene Antagonists/ use ppez 2974 14 exp leukotriene receptor blocking agent/ use emczd 18407 15 exp montelukast/ use emczd 8443 16 (montelukast or singulair).mp. 10840 17 exp Proton Pump Inhibitors/ use ppez 16773 18 exp proton pump inhibitor/ use emczd 68622 19 (proton pump inhibitor* or PPI* or omeprazole or Prilosec or Yosprala or 88470 lansoprazole or Prevacid or dexlansoprazole or Dexilent or rabeprazole or Aciphex or pantoprazole or Protonix or esomeprazole or Nexium or Vimovo or Zegerid).ti,ab. 20 exp monoclonal antibody/ use emczd 480638 21 exp Antibodies, Monoclonal, Humanized/ use ppez 39896 22 exp Interleukin-5/ai [Antagonists & Inhibitors] 293 23 exp Receptors, Interleukin-5/ai [Antagonists & Inhibitors] 31 47

24 exp interleukin 5/ use emczd 25332 25 (anti-il-5* or Interleukin-5 or Mepolizumab or nucala or Reslizumab or Cinqair6850 or Benralizumab or fasenra).ti,ab. 26 exp Interleukin-13/ai [Antagonists & Inhibitors] 219 27 interleukin 13/ use emczd 20432 28 exp Interleukin-4/ai [Antagonists & Inhibitors] 501 29 exp interleukin 4/ use emczd 71100 30 (Dupilumab or Dupixent or anti-il-13* or anti-il-4* or Interleukin-13 or 26804 Interleukin-4).ti,ab. 31 exp Immunoglobulin E/ai [Antagonists & Inhibitors] 20 32 immunoglobulin e/ use emczd or immunoglobulin e antibody/ use emczd 80681 33 (anti-IGE or Omalizumab or Xolair).mp. 14248 34 exp Dilatation/ use ppez 10447 35 exp esophagus dilatation/ use emczd 3265 36 (dilation or dilatation).ti,ab. 210743 37 exp Immunologic Factors/ use ppez 127460 5 38 exp immunomodulating agent/ use emczd 104544 0 39 immunomodulator*.ti,ab. 93101 40 exp Mercaptopurine/ 46362 41 (6-Mercaptopurine or 6-MP or Purinethol or Purixan or azothioprine or 46642 Azathioprine or Azasan or Imuran).ti,ab. 42 exp Tumor Necrosis Factor-alpha/ai [Antagonists & Inhibitors] 14168 43 tumor necrosis factor inhibitor/ use emczd 10354 44 (Infliximab or remicade or adalimumab or humira or golimumab or simponi45437 or vedolizumab or entyvio or tofacitinib or Xeljanz or Jakvinus).ti,ab. 45 exp Cromolyn Sodium/ use ppez 4044 46 exp cromoglycate disodium/ use emczd 15471 47 cromolyn.ti,ab. 2968 48 Food Hypersensitivity/di [Diagnosis] 6271 49 exp Protein Array Analysis/ use ppez 7467 48

50 exp IMMUNOASSAY/ use ppez 475054 51 exp allergy test/ use emczd or exp food allergy/ use emczd 35286 52 (Component Resolved Diagnostic Testing or crd).ti,ab. 6533 53 (test* adj2 (allerg* or hypersensitivit*)).ti,ab. 17189 54 or/10-53 463182 7 55 9 and 54 2875 56 (Randomized Controlled Trial or Pragmatic Clinical Trial or meta-analysis).pt.550035 57 exp Randomized Controlled Trials as Topic/ 263799 58 "Randomized Controlled Trial (topic)"/ 145282 59 Randomized Controlled Trial/ or Randomization/ or Random Allocation/ or142074 Double-Blind Method/ or Double Blind Procedure/ or Double-Blind Studies/8 or Single-Blind Method/ or Single Blind Procedure/ or Single-Blind Studies/ or Placebos/ or Placebo/ 60 meta-analysis/ or systematic review/ or meta-analysis as topic/ or "meta 416180 analysis (topic)"/ or "systematic review (topic)"/ or exp technology assessment, biomedical/ 61 (random* or sham or placebo*).ti,ab,hw,kf,kw. 310035 7 62 ((singl* or doubl*) adj (blind* or dumm* or mask*)).ti,ab,hw,kf,kw. 491550 63 ((tripl* or trebl*) adj (blind* or dumm* or mask*)).ti,ab,hw,kf,kw. 1696 64 ((systematic* adj3 (review* or overview*)) or (methodologic* adj3 (review*303565 or overview*))).ti,ab,kf,kw. 65 ((quantitative adj3 (review* or overview* or synthes*)) or (research adj3 19173 (integrati* or overview*))).ti,ab,kf,kw. 66 ((integrative adj3 (review* or overview*)) or (collaborative adj3 (review* or49299 overview*)) or (pool* adj3 analy*)).ti,ab,kf,kw. 67 (data synthes* or data extraction* or data abstraction* or handsearch* or 151747 hand search or mantel haenszel or peto or der simonian or dersimonian or fixed effect* or latin square* or met analy* or metanaly* or technology assessment* or HTA or HTAs or technology overview* or technology appraisal* or meta regression* or metaregression* or outcomes research or relative effectiveness).ti,ab,kf,kw. 68 (cochrane or (health adj2 technology assessment) or evidence report).jw.44072 69 (meta-analy* or metaanaly* or systematic review* or biomedical technology594805 assessment* or bio-medical technology assessment*).mp,hw. 49

70 (medline or cochrane or pubmed or medlars or embase or cinahl).ti,ab,hw.387579 71 (meta-analysis or systematic review).md. 0 72 (comparative adj3 (efficacy or effectiveness)).ti,ab,kf,kw. 27349 73 ((indirect or indirect treatment or mixed-treatment) adj 4938 comparison*).ti,ab,kf,kw. 74 or/56-73 374998 1 75 55 and 74 420 76 exp diet therapy/ 380595 77 (diet* adj2 (6 food or six food or elimination or elemental)).ti,ab. 5676 78 76 or 77 385128 79 9 and 78 875 80 75 or 79 1166 81 animals/ not (humans/ and animals/) 576852 0 82 80 not 81 1163 83 limit 82 to (case reports or comment or editorial or letter or conference 418 abstract or note) [Limit not valid in Ovid MEDLINE(R),Ovid MEDLINE(R) Daily Update,Ovid MEDLINE(R) In-Process,Ovid MEDLINE(R) Publisher,Embase; records were retained] 84 case report/ 421168 7 85 82 not (83 or 84) 722 86 remove duplicates from 85 496 1186 1187 Wiley Cochrane 1188 ID Search Hits 1189 #1 MeSH descriptor: [Eosinophilic Esophagitis] explode all trees 26 1190 #2 (eosinophilic near/2 (esophagitis or oesophagitis)):ti,ab 126 1191 #3 #1 or #2 128 1192 #4 MeSH descriptor: [Eosinophilia] explode all trees 240 1193 #5 MeSH descriptor: [Esophagitis] explode all trees 680 1194 #6 MeSH descriptor: [Esophagus] explode all trees 1282 1195 #7 #4 and (#5 or #6) 36 1196 #8 eoe:ti,ab 86 1197 #9 #3 or #7 or #8 131 1198 #10 MeSH descriptor: [Adrenal Cortex Hormones] explode all trees 13563 50

1199 #11 (corticosteroid* or steroid* or prednisone or budesonide or 1200 fluticasone):ti,ab 35079 1201 #12 MeSH descriptor: [Leukotriene Antagonists] explode all trees 439 1202 #13 (montelukast or singulair):ti,ab 1269 1203 #14 MeSH descriptor: [Proton Pump Inhibitors] explode all trees 1302 1204 #15 (proton pump inhibitor* or PPI* or omeprazole or Prilosec or Yosprala or 1205 lansoprazole or Prevacid or dexlansoprazole or Dexilent or rabeprazole or Aciphex or 1206 pantoprazole or Protonix or esomeprazole or Nexium or Vimovo or Zegerid):ti,ab 1207 6841 1208 #16 MeSH descriptor: [Antibodies, Monoclonal, Humanized] explode all trees 1209 3971 1210 #17 MeSH descriptor: [Interleukin-5] explode all trees 159 1211 #18 MeSH descriptor: [Receptors, Interleukin-5] explode all trees 9 1212 #19 (anti-il-5* or Interleukin-5 or Mepolizumab or nucala or Reslizumab or 1213 Cinqair or Benralizumab or fasenra):ti,ab 806 1214 #20 MeSH descriptor: [Interleukin-13] explode all trees 81 1215 #21 MeSH descriptor: [Interleukin-4] explode all trees 332 1216 #22 (Dupilumab or Dupixent or anti-il-13* or anti-il-4* or Interleukin-13 or 1217 Interleukin-4):ti,ab 873 1218 #23 MeSH descriptor: [Immunoglobulin E] explode all trees 1287 1219 #24 (anti-IGE or Omalizumab or Xolair):ti,ab 704 1220 #25 MeSH descriptor: [Dilatation] explode all trees 420 1221 #26 (dilation or dilatation):ti,ab 6065 1222 #27 MeSH descriptor: [Immunologic Factors] explode all trees 8197 1223 #28 immunomodulator*:ti,ab 2491 1224 #29 MeSH descriptor: [Mercaptopurine] explode all trees 1304 1225 #30 (6-Mercaptopurine or 6-MP or Purinethol or Purixan or azothioprine or 1226 Azathioprine or Azasan or Imuran):ti,ab 2101 1227 #31 MeSH descriptor: [Tumor Necrosis Factor-alpha] explode all trees 1228 3094 1229 #32 (Infliximab or remicade or adalimumab or humira or golimumab or simponi 1230 or vedolizumab or entyvio or tofacitinib or Xeljanz or Jakvinus):ti,ab 3531 1231 #33 MeSH descriptor: [Cromolyn Sodium] explode all trees 705 1232 #34 cromolyn:ti,ab 332 1233 #35 MeSH descriptor: [Food Hypersensitivity] explode all trees 731 1234 #36 MeSH descriptor: [Protein Array Analysis] explode all trees 34 1235 #37 MeSH descriptor: [Immunoassay] explode all trees 4804 1236 #38 (Component Resolved Diagnostic Testing or crd):ti,ab 166 1237 #39 (test* near/2 (allerg* or hypersensitivit*)):ti,ab 756 1238 #40 MeSH descriptor: [Diet Therapy] explode all trees 5354 1239 #41 (diet* near/2 (6 food or six food or elimination or elemental)):ti,ab 306 1240 #42 #10 or #11 or #12 or #13 or #14 or #15 or #16 or #17 or #18 or #19 or #20 1241 or #21 or #22 or #23 or #24 or #25 or #26 or #27 or #28 or #29 or #30 or #31 or #32 or 1242 #33 or #34 or #35 or #36 or #37 or #38 or #39 or #40 or #41 86910 1243 #43 #9 and #42 103 1244 51

1245 . 1246 1247 1248 1249 1250 1251 52

1252 Appendix Table 4. Additional Evidence Profile 1253 1254 Bibliography: Peterson 2010, Moawad 2013 Certainty assessment № of patients Effect

Topica Relati Absolu Certain Importan № of Risk Other Study Inconsisten Indirectne Imprecisi l ve te ty ce studie of consideratio PPI design cy ss on Steroi (95% (95% s bias ns ds CI) CI)

Failure to achieve histologic response (assessed with: < 15 Eos/hpf)

2 randomis not not serious serious a serious b none 53/88 55/88 RR 25 ⨁⨁◯ CRITICA ed trials serio (60.2% (62.5 0.96 fewer ◯ L us ) %) (0.76 per LOW to 1,000 1.22) (from 137 more to 150 fewer)

1255 CI: Confidence interval; RR: Risk ratio

1256 Explanations 1257 a. No PPI trial before steroids (so GERD is possible confounder) 1258 b. Clinical management would vary at ends of CI 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 53

1284 Figure 1. PRISMA Flow Diagram 1285 1286 1287 1288 1289 1290 Records identified through Additional records identified through other sources 1291 database searching (expert input and systematic reviews) (n = 47)* All studies that were identified before the 1292 (n = 453 ) search (and not specifically searched for) 1293

1294 1295 Records after duplicates removed 1296 (n =500) 1297 1298 1299 1300 Records screened Records excluded 1301 (n =500 ) (n =373 ) 1302 1303 1304 1305 Full-text articles assessed Full-text articles excluded, 1306 for eligibility with reasons 1307 (n =127 ) (n =40 ) 1308 Most excluded because 1309 systematic reviews, others 1310 Studies included in failed to meet 1311 qualitative/quantitative inclusion/exclusion 1312 synthesis 1313 (n = 87 ) 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 54

1325 Figure 2. Forest Plots for each PICO question 1326 1327 #1 Should proton pump inhibitors be used in patients with esophageal 1328 eosinophilia? 1329

1330 1331 1332 1333 1334 1335 #2 Should topical glucocorticosteroids be used in patients with EoE? 1336 1337

1338 1339 1340 1341 55

1342 Adverse events in topical steroids compared to placebo 1343

1344 1345 1346 1347 #3 Should systemic glucocorticosteroids be used in patients with EoE? 1348 1349 Forest plot not shown (only 1 study) 1350 1351 1352 #4 Should an elemental diet be used in patients with EoE? 1353

1354 1355 1356 #5 Should an empiric food elimination diet be used in patients with EoE? 1357 1358 6 food elimination

1359 56

1360 1361 4 food elimination

1362 1363 2 food elimination

1364 1365 Single food elimination

1366 1367 1368 1369 #6 Should allergy-based testing be used for the purpose of identifying food 1370 triggers in patients with EoE?

1371 1372 57

1373 Studies including patch testing

1374 1375 Studies not using patch testing

1376 1377 1378 #7 Should maintenance therapy be recommended in patients with EoE? 1379 1380 Forest Plot not shown (only 1 study) 1381 1382 #8 Should esophageal dilation be used in patients with EoE? 1383 1384 1385 1386 Clinical Improvement: 1387

1388 1389 58

1390 1391 1392 1393 Mortality 1394

1395 1396 1397 1398 Perforation 1399

1400 1401 1402 1403 Hospitalizations 59

1404 1405 1406 60

1407 1408 1409 Hemorrhage

1410 1411 1412 1413 1414 #9 Should anti-IL-5 therapy be used in patients with EoE? 1415

1416 1417 1418 1419 #10 Should anti-IL-13 therapy be used in patients with EoE? 1420 1421 Forest plot not shown (only 1 study) 1422 1423 #11 Should anti-IgE therapy be used in patients with EoE? 1424 1425 Forest plot not shown (only 1 study) 1426 1427 #12 Should montelukast be used in patients with EoE? 1428 1429 Forest plot not shown (only 1 study) 1430 1431 #13 Should cromolyn be used in patients with EoE? 1432 1433 Forest plot not shown (only 1 study) 1434 1435 #14 Should anti-TNFs be used in pts with EoE? 1436 1437 Forest plot not shown (only 1 study) 61

1438 1439 #15 Should immunomodulators be used in the treatment of EoE? 1440 1441 Forest plot not shown (only 1 study) 1442 1443 REFERENCES 1444 1445 1. Dellon ES, Hirano I. Epidemiology and Natural History of Eosinophilic Esophagitis. 1446 Gastroenterology. 2018;154(2):319-32.e3. 1447 2. Mukkada V, Falk GW, Eichinger CS, King D, Todorova L, Shaheen NJ. Health-Related 1448 Quality of Life and Costs Associated With Eosinophilic Esophagitis: A Systematic Review. Clin 1449 Gastroenterol Hepatol. 2018;16(4):495-503.e8. 1450 3. Jensen ET, Kappelman MD, Martin CF, Dellon ES. Health-care utilization, costs, and the 1451 burden of disease related to eosinophilic esophagitis in the United States. The American journal 1452 of gastroenterology. 2015;110(5):626-32. 1453 4. Dellon ES, Gonsalves N, Hirano I, Furuta GT, Liacouras CA, Katzka DA. ACG clinical 1454 guideline: Evidenced based approach to the diagnosis and management of esophageal 1455 eosinophilia and eosinophilic esophagitis (EoE). The American journal of gastroenterology. 1456 2013;108(5):679-92; quiz 93. 1457 5. Furuta GT, Liacouras CA, Collins MH, Gupta SK, Justinich C, Putnam PE, et al. 1458 Eosinophilic esophagitis in children and adults: a systematic review and consensus 1459 recommendations for diagnosis and treatment. Gastroenterology. 2007;133(4):1342-63. 1460 6. Liacouras CA, Furuta GT, Hirano I, Atkins D, Attwood SE, Bonis PA, et al. Eosinophilic 1461 esophagitis: Updated consensus recommendations for children and adults. Journal of Allergy 1462 and Clinical Immunology. 2011;128(1):3-20. 1463 7. Molina-Infante J, Bredenoord AJ, Cheng E, Dellon ES, Furuta GT, Gupta SK, et al. 1464 Proton pump inhibitor-responsive oesophageal eosinophilia: an entity challenging current 1465 diagnostic criteria for eosinophilic oesophagitis. Gut. 2016;65(3):524-31. 1466 8. Lucendo AJ, Molina-Infante J, Arias A, von Arnim U, Bredenoord AJ, Bussmann C, et al. 1467 Guidelines on eosinophilic esophagitis: evidence-based statements and recommendations for 1468 diagnosis and management in children and adults. United European Gastroenterology Journal. 1469 2017;5(3):335-58. 1470 9. Schunemann HJ, Oxman AD, Brozek J, Glasziou P, Jaeschke R, Vist GE, et al. Grading 1471 quality of evidence and strength of recommendations for diagnostic tests and strategies. Bmj. 1472 2008;336(7653):1106-10. 1473 10. Reed CC, Wolf WA, Cotton CC, Rusin S, Perjar I, Hollyfield J, et al. Optimal Histologic 1474 Cutpoints for Treatment Response in Patients With Eosinophilic Esophagitis: Analysis of Data 1475 From a Prospective Cohort Study. Clinical Gastroenterology and Hepatology. 2018;16(2):226- 1476 33.e2. 1477 11. Wolf WA, Cotton CC, Green DJ, Hughes JT, Woosley JT, Shaheen NJ, et al. Evaluation 1478 of Histologic Cutpoints for Treatment Response in Eosinophilic Esophagitis. J Gastroenterol 1479 Hepatol Res. 2015;4(10):1780-7. 1480 12. Warners MJ, Oude Nijhuis RAB, de Wijkerslooth LRH, Smout A, Bredenoord AJ. The 1481 natural course of eosinophilic esophagitis and long-term consequences of undiagnosed disease 1482 in a large cohort. The American journal of gastroenterology. 2018;113(6):836-44. 1483 13. Schoepfer AM, Safroneeva E, Bussmann C, Kuchen T, Portmann S, Simon HU, et al. 1484 Delay in diagnosis of eosinophilic esophagitis increases risk for stricture formation in a time- 1485 dependent manner. Gastroenterology. 2013;145(6):1230-6.e1-2. 62

1486 14. Lipka S, Kumar A, Richter JE. Impact of Diagnostic Delay and Other Risk Factors on 1487 Eosinophilic Esophagitis Phenotype and Esophageal Diameter. J Clin Gastroenterol. 1488 2016;50(2):134-40. 1489 15. Dellon ES, Kim HP, Sperry SL, Rybnicek DA, Woosley JT, Shaheen NJ. A phenotypic 1490 analysis shows that eosinophilic esophagitis is a progressive fibrostenotic disease. Gastrointest 1491 Endosc. 2014;79(4):577-85.e4. 1492 16. Gupta S, Vitanza J, Collins M. Efficacy and safety of oral budesonide suspension in 1493 pediatric patients with eosinophilic esophagitis. Clinical gastroenterology and hepatology 1494 [Internet]. 2015; 13(1):[66-76.e3 pp.]. Available from: http://cochranelibrary- 1495 wiley.com/o/cochrane/clcentral/articles/051/CN-01116051/frame.html. 1496 17. Reed CC, Wolf WA, Cotton CC, Dellon ES. A visual analogue scale and a Likert scale 1497 are simple and responsive tools for assessing dysphagia in eosinophilic oesophagitis. 1498 Alimentary Pharmacology and Therapeutics. 2017;45(11):1443-8. 1499 18. Safroneeva E, Straumann A, Coslovsky M, Zwahlen M, Kuehni CE, Panczak R, et al. 1500 Symptoms Have Modest Accuracy in Detecting Endoscopic and Histologic Remission in Adults 1501 With Eosinophilic Esophagitis. Gastroenterology. 2016;150(3):581-90.e4. 1502 19. Alexander J, Jung K, Arora A, Enders F, Katzka D, Kephardt G, et al. Swallowed 1503 fluticasone improves histologic but not symptomatic response of adults with eosinophilic 1504 esophagitis. Clinical gastroenterology and hepatology [Internet]. 2012; 10(7):[742-9.e1 pp.]. 1505 Available from: http://cochranelibrary-wiley.com/o/cochrane/clcentral/articles/442/CN- 1506 00842442/frame.html. 1507 20. Butz B, Wen T, Gleich G, Furuta G, Spergel J, King E, et al. Efficacy, dose reduction, 1508 and resistance to high-dose fluticasone in patients with eosinophilic esophagitis. 1509 Gastroenterology [Internet]. 2014; 147(2):[324-33.e5 pp.]. Available from: http://cochranelibrary- 1510 wiley.com/o/cochrane/clcentral/articles/917/CN-01037917/frame.html. 1511 21. Dellon E, Katzka D, Collins M, Hamdani M, Gupta S, Hirano I. Budesonide Oral 1512 Suspension Improves Symptomatic, Endoscopic, and Histologic Parameters Compared 1513 With Placebo in Patients With Eosinophilic Esophagitis. Gastroenterology [Internet]. 2017; 1514 152(4):[776-86.e5 pp.]. Available from: http://cochranelibrary- 1515 wiley.com/o/cochrane/clcentral/articles/078/CN-01382078/frame.html. 1516 22. Dohil R, Newbury R, Fox L, Bastian J, Aceves S. Oral viscous budesonide is effective in 1517 children with eosinophilic esophagitis in a randomized, placebo-controlled trial. 1518 Gastroenterology. 2010;139(2):418-29.e1. 1519 23. Gupta SK, Vitanza JM, Collins MH. Efficacy and safety of oral budesonide suspension in 1520 pediatric patients with eosinophilic esophagitis. Clin Gastroenterol Hepatol. 2015;13(1):66- 1521 76.e3. 1522 24. Konikoff M, Noel R, Blanchard C, Kirby C, Jameson S, Buckmeier B, et al. A 1523 randomized, double-blind, placebo-controlled trial of fluticasone propionate for pediatric 1524 eosinophilic esophagitis. Gastroenterology [Internet]. 2006; 131(5):[1381-91 pp.]. Available 1525 from: http://cochranelibrary-wiley.com/o/cochrane/clcentral/articles/657/CN- 1526 00573657/frame.html. 1527 25. Miehlke S, Hruz P, Vieth M, Bussmann C, Arnim U, Bajbouj M, et al. A randomised, 1528 double-blind trial comparing budesonide formulations and dosages for short-term treatment of 1529 eosinophilic oesophagitis. Gut [Internet]. 2016; 65(3):[390-9 pp.]. Available from: 1530 http://cochranelibrary-wiley.com/o/cochrane/clcentral/articles/980/CN-01136980/frame.html. 1531 26. Straumann A, Conus S, Degen L, Felder S, Kummer M, Engel H, et al. Budesonide is 1532 effective in adolescent and adult patients with active eosinophilic esophagitis. Gastroenterology 1533 [Internet]. 2010; 139(5):[1526-37, 37.e1 pp.]. Available from: http://cochranelibrary- 1534 wiley.com/o/cochrane/clcentral/articles/029/CN-00770029/frame.html. 1535 27. Garrean C, Hirano I. Eosinophilic esophagitis: Pathophysiology and optimal 1536 management. Current Gastroenterology Reports. 2009;11(3):175-81. 63

1537 28. Peterson K, Thomas K, Hilden K, Emerson L, Wills J, Fang J. Comparison of 1538 esomeprazole to aerosolized, swallowed fluticasone for eosinophilic esophagitis. Digestive 1539 diseases and sciences [Internet]. 2010; 55(5):[1313-9 pp.]. Available from: 1540 http://cochranelibrary-wiley.com/o/cochrane/clcentral/articles/867/CN-00749867/frame.html. 1541 29. Molina-Infante J, Ferrando-Lamana L, Ripoll C, Hernandez-Alonso M, Mateos JM, 1542 Fernandez-Bermejo M, et al. Esophageal eosinophilic infiltration responds to proton pump 1543 inhibition in most adults. Clin Gastroenterol Hepatol. 2011;9(2):110-7. 1544 30. Abe Y, Iijima K, Ohara S, Koike T, Ara N, Uno K, et al. A Japanese case series of 12 1545 patients with esophageal eosinophilia. J Gastroenterol. 2011;46(1):25-30. 1546 31. Fujiwara Y, Sugawa T, Tanaka F, Tatsuwaki H, Okuyama M, Hayakawa T, et al. A 1547 multicenter study on the prevalence of eosinophilic esophagitis and PPI-responsive esophageal 1548 eosinophilic infiltration. Internal medicine (Tokyo, Japan). 2012;51(23):3235-9. 1549 32. Francis D, Foxx-Orenstein A, Arora A, Smyrk T, Jensen K, Nord S, et al. Results of 1550 ambulatory pH monitoring do not reliably predict response to therapy in patients with 1551 eosinophilic oesophagitis. Alimentary pharmacology & therapeutics [Internet]. 2012; 35(2):[300- 1552 7 pp.]. Available from: http://cochranelibrary-wiley.com/o/cochrane/clcentral/articles/430/CN- 1553 00901430/frame.html. 1554 33. Vazquez-Elizondo G, Ngamruengphong S, Khrisna M, Devault KR, Talley NJ, Achem 1555 SR. The outcome of patients with oesophageal eosinophilic infiltration after an eight-week trial 1556 of a proton pump inhibitor. Aliment Pharmacol Ther. 2013;38(10):1312-9. 1557 34. Moawad F, Veerappan G, Dias J, Baker T, Maydonovitch C, Wong R. Randomized 1558 controlled trial comparing aerosolized swallowed fluticasone to esomeprazole for esophageal 1559 eosinophilia. American journal of gastroenterology [Internet]. 2013; 108(3):[366-72 pp.]. 1560 Available from: http://cochranelibrary-wiley.com/o/cochrane/clcentral/articles/389/CN- 1561 00877389/frame.html. 1562 35. Lee JH, Kim MJ, Kim JH, Youn YH, Kim N, Bak YT, et al. Clinical analysis of primary 1563 eosinophilic esophagitis. Journal of neurogastroenterology and motility. 2013;19(2):204-9. 1564 36. Dellon ES, Speck O, Woodward K, Gebhart JH, Madanick RD, Levinson S, et al. Clinical 1565 and endoscopic characteristics do not reliably differentiate PPI-responsive esophageal 1566 eosinophilia and eosinophilic esophagitis in patients undergoing upper endoscopy: a 1567 prospective cohort study. The American journal of gastroenterology. 2013;108(12):1854-60. 1568 37. Mangla S SG, Hornick JL, et al. . Clinical predictors of response to proton pump 1569 inhibitors in patients with esophageal eosinophilia. Gastroenterology. 2013;;144 (Suppl 1570 1)::S495-6 (Su1866). 1571 38. Molina-Infante J, Arias A, Barrio J, Rodriguez-Sanchez J, Sanchez-Cazalilla M, Lucendo 1572 AJ. Four-food group elimination diet for adult eosinophilic esophagitis: A prospective multicenter 1573 study. Journal of Allergy and Clinical Immunology. 2014;134(5):1093-9.e1. 1574 39. van Rhijn BD, Weijenborg PW, Verheij J, van den Bergh Weerman MA, Verseijden C, 1575 van den Wijngaard RM, et al. Proton pump inhibitors partially restore mucosal integrity in 1576 patients with proton pump inhibitor-responsive esophageal eosinophilia but not eosinophilic 1577 esophagitis. Clin Gastroenterol Hepatol. 2014;12(11):1815-23.e2. 1578 40. Gomez-Torrijos E, Garcia-Rodriguez R, Castro-Jimenez A, Rodriguez-Sanchez J, 1579 Mendez Diaz Y, Molina-Infante J. The efficacy of step-down therapy in adult patients with proton 1580 pump inhibitor-responsive oesophageal eosinophilia. Aliment Pharmacol Ther. 2016;43(4):534- 1581 40. 1582 41. Jiao D, Ishimura N, Maruyama R, Ishikawa N, Nagase M, Oshima N, et al. Similarities 1583 and differences among eosinophilic esophagitis, proton-pump inhibitor-responsive esophageal 1584 eosinophilia, and reflux esophagitis: comparisons of clinical, endoscopic, and histopathological 1585 findings in Japanese patients. J Gastroenterol. 2017;52(2):203-10. 1586 42. Savarino EV, Tolone S, Bartolo O, de Cassan C, Caccaro R, Galeazzi F, et al. The 1587 GerdQ questionnaire and high resolution manometry support the hypothesis that proton pump 64

1588 inhibitor-responsive oesophageal eosinophilia is a GERD-related phenomenon. Aliment 1589 Pharmacol Ther. 2016;44(5):522-30. 1590 43. Philpott H, Nandurkar S, Royce S, Thien F, Gibson P. A prospective open clinical trial of 1591 a proton pump inhibitor, elimination diet and/or budesonide for eosinophilic oesophagitis. 1592 Alimentary pharmacology & therapeutics [Internet]. 2016; 43(9):[985-93 pp.]. Available from: 1593 http://cochranelibrary-wiley.com/o/cochrane/clcentral/articles/852/CN-01140852/frame.html. 1594 44. Sayej WN, Patel R, Baker RD, Tron E, Baker SS. Treatment with high-dose proton pump 1595 inhibitors helps distinguish eosinophilic esophagitis from noneosinophilic esophagitis. J Pediatr 1596 Gastroenterol Nutr. 2009;49(4):393-9. 1597 45. Dranove JE, Horn DS, Davis MA, Kernek KM, Gupta SK. Predictors of response to 1598 proton pump inhibitor therapy among children with significant esophageal eosinophilia. The 1599 Journal of pediatrics. 2009;154(1):96-100. 1600 46. Schroeder S, Capocelli KE, Masterson JC, Harris R, Protheroe C, Lee JJ, et al. Effect of 1601 proton pump inhibitor on esophageal eosinophilia. J Pediatr Gastroenterol Nutr. 2013;56(2):166- 1602 72. 1603 47. Rea F, Caldaro T, Tambucci R, Romeo EF, Caloisi C, Torroni F, et al. Eosinophilic 1604 esophagitis: Is it also a surgical disease? Journal of Pediatric Surgery. 2013;48(2):304-8. 1605 48. Dhaliwal J, Tobias V, Sugo E, Varjavandi V, Lemberg D, Day A, et al. Eosinophilic 1606 esophagitis in children with esophageal atresia. Dis Esophagus. 2014;27(4):340-7. 1607 49. Gutierrez-Junquera C, Fernandez-Fernandez S, Cilleruelo ML, Rayo A, Echeverria L, 1608 Quevedo S, et al. High Prevalence of Response to Proton-pump Inhibitor Treatment in Children 1609 With Esophageal Eosinophilia. J Pediatr Gastroenterol Nutr. 2016;62(5):704-10. 1610 50. Straumann A, Conus S, Degen L, Frei C, Bussmann C, Beglinger C, et al. Long-Term 1611 Budesonide Maintenance Treatment Is Partially Effective for Patients With Eosinophilic 1612 Esophagitis. Clinical Gastroenterology and Hepatology. 2011;9(5):400-9.e1. 1613 51. Schaefer E, Fitzgerald J, Molleston J, Croffie J, Pfefferkorn M, Corkins M, et al. 1614 Comparison of oral prednisone and topical fluticasone in the treatment of eosinophilic 1615 esophagitis: a randomized trial in children. Clinical gastroenterology and hepatology [Internet]. 1616 2008; 6(2):[165-73 pp.]. Available from: http://cochranelibrary- 1617 wiley.com/o/cochrane/clcentral/articles/409/CN-00622409/frame.html. 1618 52. Moawad FJ, Veerappan GR, Dias JA, Baker TP, Maydonovitch CL, Wong RKH. 1619 Randomized controlled trial comparing aerosolized swallowed fluticasone to esomeprazole for 1620 esophageal eosinophilia. American Journal of Gastroenterology. 2013;108(3):366-72. 1621 53. Dellon E, Sheikh A, Speck O, Woodward K, Whitlow A, Hores J, et al. Viscous topical is 1622 more effective than nebulized steroid therapy for patients with eosinophilic esophagitis. 1623 Gastroenterology [Internet]. 2012; 143(2):[321-4.e1 pp.]. Available from: http://cochranelibrary- 1624 wiley.com/o/cochrane/clcentral/articles/750/CN-00832750/frame.html. 1625 54. Bhardwaj N, Ishmael F, Lehman E, Bethards D, Ruggiero F, Ghaffari G. Effect of topical 1626 beclomethasone on inflammatory markers in adults with eosinophilic esophagitis: A pilot study. 1627 Allergy & Rhinology. 2017;8(2):85-94. 1628 55. Philpott H, Dougherty MK, Reed CC, Caldwell M, Kirk D, Torpy DJ, et al. Systematic 1629 review: adrenal insufficiency secondary to swallowed topical corticosteroids in eosinophilic 1630 oesophagitis. Aliment Pharmacol Ther. 2018;47(8):1071-8. 1631 56. Chang A, Robison R, Cai M, Singh AM. Natural History of Food-Triggered Atopic 1632 Dermatitis and Development of Immediate Reactions in Children. The journal of allergy and 1633 clinical immunology In practice. 2016;4(2):229-36.e1. 1634 57. Du Toit G, Roberts G, Sayre PH, Bahnson HT, Radulovic S, Santos AF, et al. 1635 Randomized trial of peanut consumption in infants at risk for peanut allergy. The New England 1636 journal of medicine. 2015;372(9):803-13. 65

1637 58. Kelly KJ, Lazenby AJ, Rowe PC, Yardley JH, Perman JA, Sampson HA. Eosinophilic 1638 esophagitis attributed to gastroesophageal reflux: improvement with an amino acid-based 1639 formula. Gastroenterology. 1995;109(5):1503-12. 1640 59. Liacouras CA, Spergel JM, Ruchelli E, Verma R, Mascarenhas M, Semeao E, et al. 1641 Eosinophilic esophagitis: A 10-year experience in 381 children. Clinical Gastroenterology and 1642 Hepatology. 2005;3(12):1198-206. 1643 60. Henderson CJ, Abonia JP, King EC, Putnam PE, Collins MH, Franciosi JP, et al. 1644 Comparative dietary therapy effectiveness in remission of pediatric eosinophilic esophagitis. 1645 The Journal of allergy and clinical immunology. 2012;129(6):1570-8. 1646 61. Peterson KA, Byrne KR, Vinson LA, Ying J, Boynton KK, Fang JC, et al. Elemental diet 1647 induces histologic response in adult eosinophilic esophagitis. American Journal of 1648 Gastroenterology. 2013;108(5):759-66. 1649 62. Leung J, Mehrzad R, Hundal NV, Alejos A, Hesterberg PE, Katz AJ, et al. Longitudinal 1650 Perspective on Managing Refractory Eosinophilic Esophagitis. The journal of allergy and clinical 1651 immunology In practice. 2015;3(6):951-6. 1652 63. Warners MJ, Vlieg-Boerstra BJ, Verheij J, van Rhijn BD, Van Ampting MTJ, Harthoorn 1653 LF, et al. Elemental diet decreases inflammation and improves symptoms in adult eosinophilic 1654 oesophagitis patients. Alimentary Pharmacology and Therapeutics. 2017;45(6):777-87. 1655 64. Kagalwalla AF, Sentongo TA, Ritz S, Hess T, Nelson SP, Emerick KM, et al. Effect of 1656 six-food elimination diet on clinical and histologic outcomes in eosinophilic esophagitis. Clinical 1657 Gastroenterology & Hepatology. 2006;4(9):1097-102. 1658 65. Gonsalves N, Yang GY, Doerfler B, Ritz S, Ditto AM, Hirano I. Elimination diet effectively 1659 treats eosinophilic esophagitis in adults; Food reintroduction identifies causative factors. 1660 Gastroenterology. 2012;142(7):1451-9.e1. 1661 66. Henderson CJ, Abonia JP, King EC, Putnam PE, Collins MH, Franciosi JP, et al. 1662 Comparative dietary therapy effectiveness in remission of pediatric eosinophilic esophagitis. 1663 Journal of Allergy & Clinical Immunology. 2012;129(6):1570-8. 1664 67. Lucendo A, Arias Á, González-Cervera J, Yagüe-Compadre J, Guagnozzi D, Angueira 1665 T, et al. Empiric 6-food elimination diet induced and maintained prolonged remission in patients 1666 with adult eosinophilic esophagitis: a prospective study on the food cause of the disease. 1667 Journal of allergy and clinical immunology [Internet]. 2013; 131(3):[797-804 pp.]. Available from: 1668 http://cochranelibrary-wiley.com/o/cochrane/clcentral/articles/622/CN-00965622/frame.html. 1669 68. Colson D, Kalach N, Soulaines P, Vannerom Y, Campeotto F, Talbotec C, et al. The 1670 Impact of Dietary Therapy on Clinical and Biologic Parameters of Pediatric Patients with 1671 Eosinophilic Esophagitis. Journal of Allergy and Clinical Immunology: In Practice. 1672 2014;2(5):587-93. 1673 69. Rodriguez-Sanchez J, Gomez Torrijos E, Lopez Viedma B, De La Santa Belda E, Martin 1674 Davila F, Garcia Rodriguez C, et al. Efficacy of IgE-targeted vs empiric six-food elimination diets 1675 for adult eosinophilic oesophagitis. Allergy: European Journal of Allergy and Clinical 1676 Immunology. 2014;69(7):936-42. 1677 70. Philpott H, Nandurkar S, Royce SG, Thien F, Gibson PR. A prospective open clinical trial 1678 of a proton pump inhibitor, elimination diet and/or budesonide for eosinophilic oesophagitis. 1679 Alimentary Pharmacology and Therapeutics. 2016;43(9):985-93. 1680 71. Molina-Infante J, Arias A, Alcedo J, Garcia-Romero R, Casabona-Frances S, Prieto- 1681 Garcia A, et al. Step-up empiric elimination diet for pediatric and adult eosinophilic esophagitis: 1682 The 2-4-6 study. Journal of Allergy and Clinical Immunology. 2018;141(4):1365-72. 1683 72. Reed CC, Fan C, Koutlas NT, Shaheen NJ, Dellon ES. Food elimination diets are 1684 effective for long-term treatment of adults with eosinophilic oesophagitis. Alimentary 1685 Pharmacology and Therapeutics. 2017;46(9):836-44. 66

1686 73. Homan M, Blagus R, Koren Jeverica A, Orel R. Pediatric eosinophilic esophagitis in 1687 Slovenia: Data from a retrospective 2005-2012 epidemiological study. Journal of Pediatric 1688 Gastroenterology and Nutrition. 2015;61(3):313-8. 1689 74. Kagalwalla AF, Wechsler JB, Amsden K, Schwartz S, Makhija M, Olive A, et al. Efficacy 1690 of a 4-Food Elimination Diet for Children With Eosinophilic Esophagitis. Clinical 1691 Gastroenterology and Hepatology. 2017;15(11):1698-707.e7. 1692 75. Reed CC, Safta AM, Qasem S, Angie Almond M, Dellon ES, Jensen ET. Combined and 1693 Alternating Topical Steroids and Food Elimination Diet for the Treatment of Eosinophilic 1694 Esophagitis. Digestive Diseases and Sciences. 2018:1-8. 1695 76. Kagalwalla AF, Amsden K, Shah A, Ritz S, Manuel-Rubio M, Dunne K, et al. Cow's milk 1696 elimination: a novel dietary approach to treat eosinophilic esophagitis. J Pediatr Gastroenterol 1697 Nutr. 2012;55(6):711-6. 1698 77. Kruszewski PG, Russo JM, Franciosi JP, Varni JW, Platts-Mills TAE, Erwin EA. 1699 Prospective, comparative effectiveness trial of cow's milk elimination and swallowed fluticasone 1700 for pediatric eosinophilic esophagitis. Diseases of the Esophagus. 2016;29(4):377-84. 1701 78. Quaglietta L, Coccorullo P, Miele E, Pascarella F, Troncone R, Staiano A. Eosinophilic 1702 oesophagitis and coeliac disease: is there an association? Aliment Pharmacol Ther. 1703 2007;26(3):487-93. 1704 79. Rizo Pascual JM, De La Hoz Caballer B, Redondo Verge C, Terrados Cepeda S, Roy 1705 Arino G, Riesco Lopez JM, et al. Allergy assessment in children with eosinophilic esophagitis. 1706 Journal of Investigational Allergology & Clinical Immunology. 2011;21(1):59-65. 1707 80. Molina-Infante J, Martin-Noguerol E, Alvarado-Arenas M, Porcel-Carreno SL, Jimenez- 1708 Timon S, Hernandez-Arbeiza FJ. Selective elimination diet based on skin testing has suboptimal 1709 efficacy for adult eosinophilic esophagitis. The Journal of allergy and clinical immunology. 1710 2012;130(5):1200-2. 1711 81. Spergel JM, Brown-Whitehorn TF, Cianferoni A, Shuker M, Wang ML, Verma R, et al. 1712 Identification of causative foods in children with eosinophilic esophagitis treated with an 1713 elimination diet. Journal of Allergy and Clinical Immunology. 2012;130(2):461-7.e5. 1714 82. Al-Hussaini A, Al-Idressi E, Al-Zahrani M. The role of allergy evaluation in children with 1715 eosinophilic esophagitis. Journal of Gastroenterology. 2013;48(11):1205-12. 1716 83. Syrigou E, Angelakopoulou A, Zande M, Panagiotou I, Roma E, Pitsios C. Allergy-test- 1717 driven elimination diet is useful in children with eosinophilic esophagitis, regardless of the 1718 severity of symptoms. Pediatric Allergy and Immunology. 2015;26(4):323-9. 1719 84. Van Rhijn BD, Verheij J, Van Den Bergh Weerman MA, Verseijden C, Van Den 1720 Wijngaard RMJGJ, De Jonge WJ, et al. Histological Response to Fluticasone Propionate in 1721 Patients with Eosinophilic Esophagitis is Associated with Improved Functional Esophageal 1722 Mucosal Integrity. American Journal of Gastroenterology. 2015;110(9):1289-97. 1723 85. Constantine G, Seth N, Chokshi N, Minard CG, Guffey D, Olive AP, et al. Combination 1724 Steroid and Test-based Food Elimination for Eosinophilic Esophagitis: A Retrospective Analysis. 1725 Journal of Pediatric Gastroenterology and Nutrition. 2017;64(6):933-8. 1726 86. Andreae DA, Hanna MG, Magid MS, Malerba S, Andreae MH, Bagiella E, et al. 1727 Swallowed Fluticasone Propionate Is an Effective Long-Term Maintenance Therapy for Children 1728 With Eosinophilic Esophagitis. The American journal of gastroenterology. 2016;111(8):1187-97. 1729 87. Dellon ESKD CM, Hamdi M, et al. . Safety and efficacy of oral budesonide suspension 1730 for maintenance therapy in eosinophilic 1731 esophagitis: results from a prospective open-label study 1732 of adolescents and adults. . Gastroenterology 2016;150(Suppl 1733 1):S188 (Ab 953). 1734 88. Greuter T, Bussmann C, Safroneeva E, Schoepfer AM, Biedermann L, Vavricka SR, et 1735 al. Long-Term Treatment of Eosinophilic Esophagitis With Swallowed Topical Corticosteroids: 67

1736 Development and Evaluation of a Therapeutic Concept. The American journal of 1737 gastroenterology. 2017;112(10):1527-35. 1738 89. Eluri S, Runge TM, Hansen J, Kochar B, Reed CC, Robey BS, et al. Diminishing 1739 Effectiveness of Long-Term Maintenance Topical Steroid Therapy in PPI Non-Responsive 1740 Eosinophilic Esophagitis. Clin Transl Gastroenterol. 2017;8(6):e97. 1741 90. Alexander J, Ravi K, Enders F, Geno D, Kryzer L, Mara K, et al. Montelukast Does not 1742 Maintain Symptom Remission After Topical Steroid Therapy for Eosinophilic Esophagitis. 1743 Clinical gastroenterology and hepatology [Internet]. 2017; 15(2):[214-21.e2 pp.]. Available from: 1744 http://cochranelibrary-wiley.com/o/cochrane/clcentral/articles/040/CN-01333040/frame.html. 1745 91. Molina-Infante J, Rodriguez-Sanchez J, Martinek J, van Rhijn BD, Krajciova J, Rivas 1746 MD, et al. Long-Term Loss of Response in Proton Pump Inhibitor-Responsive Esophageal 1747 Eosinophilia Is Uncommon and Influenced by CYP2C19 Genotype and Rhinoconjunctivitis. The 1748 American journal of gastroenterology. 2015;110(11):1567-75. 1749 92. Lucendo AJ, Arias A, Gonzalez-Cervera J, Yague-Compadre JL, Guagnozzi D, Angueira 1750 T, et al. Empiric 6-food elimination diet induced and maintained prolonged remission in patients 1751 with adult eosinophilic esophagitis: A prospective study on the food cause of the disease. 1752 Journal of Allergy and Clinical Immunology. 2013;131(3):797-804. 1753 93. Philpott H, Nandurkar S, Royce SG, Thien F, Gibson PR. A prospective open clinical trial 1754 of a proton pump inhibitor, elimination diet and/or budesonide for eosinophilic oesophagitis. 1755 Aliment Pharmacol Ther. 2016;43(9):985-93. 1756 94. Dougherty M, Runge TM, Eluri S, Dellon ES. Esophageal dilation with either bougie or 1757 balloon technique as a treatment for eosinophilic esophagitis: a systematic review and meta- 1758 analysis. Gastrointestinal Endoscopy. 2017;86(4):581-91.e3. 1759 95. Moole H, Jacob K, Duvvuri A, Moole V, Dharmapuri S, Boddireddy R, et al. Role of 1760 endoscopic esophageal dilation in managing eosinophilic esophagitis. Medicine (United States). 1761 2017;96 (14) (no pagination)(e5877). 1762 96. Moawad FJ, Molina-Infante J, Lucendo AJ, Cantrell SE, Tmanova L, Douglas KM. 1763 Systematic review with meta-analysis: endoscopic dilation is highly effective and safe in children 1764 and adults with eosinophilic oesophagitis. Alimentary Pharmacology and Therapeutics. 1765 2017;46(2):96-105. 1766 97. Straumann A, Conus S, Grzonka P, Kita H, Kephart G, Bussmann C, et al. Anti- 1767 interleukin-5 antibody treatment (mepolizumab) in active eosinophilic oesophagitis: a 1768 randomised, placebo-controlled, double-blind trial. Gut [Internet]. 2010; 59(1):[21-30 pp.]. 1769 Available from: http://cochranelibrary-wiley.com/o/cochrane/clcentral/articles/524/CN- 1770 00730524/frame.html. 1771 98. Assa'ad AH, Gupta SK, Collins MH, Thomson M, Heath AT, Smith DA, et al. An antibody 1772 against IL-5 reduces numbers of esophageal intraepithelial eosinophils in children with 1773 eosinophilic esophagitis. Gastroenterology. 2011;141(5):1593-604. 1774 99. Spergel JM, Rothenberg ME, Collins MH, Furuta GT, Markowitz JE, Fuchs G, 3rd, et al. 1775 Reslizumab in children and adolescents with eosinophilic esophagitis: results of a double-blind, 1776 randomized, placebo-controlled trial. Journal of Allergy & Clinical Immunology. 2012;129(2):456- 1777 63, 63.e1-3. 1778 100. Rothenberg M, Wen T, Greenberg A, Alpan O, Enav B, Hirano I, et al. Intravenous anti- 1779 IL-13 mAb QAX576 for the treatment of eosinophilic esophagitis. Journal of allergy and clinical 1780 immunology [Internet]. 2014; 135(2):[500-7 pp.]. Available from: http://cochranelibrary- 1781 wiley.com/o/cochrane/clcentral/articles/719/CN-01072719/frame.html. 1782 101. Loizou D, Enav B, Komlodi-Pasztor E, Hider P, Kim-Chang J, Noonan L, et al. A pilot 1783 study of omalizumab in eosinophilic esophagitis. PLoS ONE. 2015;10 (3) (no 1784 pagination)(e0113483). 68

1785 102. Clayton F, Fang JC, Gleich GJ, Lucendo AJ, Olalla JM, Vinson LA, et al. Eosinophilic 1786 esophagitis in adults is associated with IgG4 and not mediated by IgE. Gastroenterology. 1787 2014;147(3):602-9. 1788 103. Attwood SE, Lewis CJ, Bronder CS, Morris CD, Armstrong GR, Whittam J. Eosinophilic 1789 oesophagitis: a novel treatment using Montelukast. Gut. 2003;52(2):181-5. 1790 104. Vanderhoof JA, Young RJ, Hanner TL, Kettlehut B. Montelukast: use in pediatric 1791 patients with eosinophilic gastrointestinal disease. J Pediatr Gastroenterol Nutr. 2003;36(2):293- 1792 4. 1793 105. Lucendo AJ, De Rezende LC, Jimenez-Contreras S, Yague-Compadre JL, Gonzalez- 1794 Cervera J, Mota-Huertas T, et al. Montelukast was inefficient in maintaining steroid-induced 1795 remission in adult eosinophilic esophagitis. Dig Dis Sci. 2011;56(12):3551-8. 1796 106. Stumphy J, Al-Zubeidi D, Guerin L, Mitros F, Rahhal R. Observations on use of 1797 montelukast in pediatric eosinophilic esophagitis: insights for the future. Dis Esophagus. 1798 2011;24(4):229-34. 1799 107. Lieberman JA, Zhang J, Whitworth J, Cavender C. A randomized, double-blinded, 1800 placebo-controlled study of the use of viscous oral cromolyn sodium for the treatment of 1801 eosinophilic esophagitis. Annals of Allergy, Asthma and Immunology. 2018;120(5):527-31. 1802 108. Straumann A, Bussmann C, Conus S, Beglinger C, Simon HU. Anti-TNF-alpha 1803 (infliximab) therapy for severe adult eosinophilic esophagitis. The Journal of allergy and clinical 1804 immunology. 2008;122(2):425-7. 1805 109. Louis-Auguste JR, Hoare J. A hard act to swallow: modern management of eosinophilic 1806 oesophagitis. Frontline gastroenterology. 2013;4(2):91-5. 1807 110. Netzer P, Gschossmann JM, Straumann A, Sendensky A, Weimann R, Schoepfer AM. 1808 Corticosteroid-dependent eosinophilic oesophagitis: azathioprine and 6-mercaptopurine can 1809 induce and maintain long-term remission. Eur J Gastroenterol Hepatol. 2007;19(10):865-9. 1810 111. Philpott H, Dellon ES. The role of maintenance therapy in eosinophilic esophagitis: who, 1811 why, and how? Journal of Gastroenterology. 2018;53(2):165-71. 1812 112. Koutlas NT, Dellon ES. Progression from an Inflammatory to a Fibrostenotic Phenotype 1813 in Eosinophilic Esophagitis. Case reports in gastroenterology. 2017;11(2):382-8. 1814 113. Safroneeva E, Coslovsky M, Kuehni CE, Zwahlen M, Haas NA, Panczak R, et al. 1815 Eosinophilic oesophagitis: Relationship of quality of life with clinical, endoscopic and histological 1816 activity. Alimentary Pharmacology and Therapeutics. 2015;42(8):1000-10. 1817 114. Spechler SJ, Genta RM, Souza RF. Thoughts on the complex relationship between 1818 gastroesophageal reflux disease and eosinophilic esophagitis. The American journal of 1819 gastroenterology. 2007;102(6):1301-6. 1820 115. Aceves SS, King E, Collins MH, Yang GY, Capocelli KE, Abonia JP, et al. Alignment of 1821 parent- and child-reported outcomes and histology in eosinophilic esophagitis across multiple 1822 CEGIR sites. The Journal of allergy and clinical immunology. 2018;142(1):130-8.e1. 1823 116. Ishimura N, Sumi S, Okada M, Mikami H, Okimoto E, Nagano N, et al. Is asymptomatic 1824 esophageal eosinophilia the same disease entity as eosinophilic esophagitis? Clin 1825 Gastroenterol Hepatol. 2018. 1826 117. Molina-Infante J, Bredenoord AJ, Cheng E, Dellon ES, Furuta GT, Gupta SK, et al. 1827 Proton pump inhibitor-responsive oesophageal eosinophilia: An entity challenging current 1828 diagnostic criteria for eosinophilic oesophagitis. Gut. 2016;65(3):521-31. 1829 1 American Gastroenterological Institute and the Joint Task Force on Formatted: Numbering: Continuous 2 Allergy-Immunology Practice Parameters Clinical Guidelines for the 3 Management of Eosinophilic Esophagitis

4 Hirano I1, Chan ES2, Rank M3, Sharaf R4, Stollman N5, Stukus D6, Wang K7 , Greenhawt M,8 Falck- 5 Ytter Y9 6 7 Collaborators: Bernstein JA10, Dinakar C11 , Golden DBK12, Khan DA13, Lieberman J, Oppenheimer J15, 8 Shaker M16, Wallace DV17, Wang J18 9 10 1 Division of Gastroenterology and Hepatology, Northwestern University Feinberg School of Medicine, 11 Chicago, Illinois; 2 Division of Allergy and Immunology, Department of Pediatrics, British Columbia Children’s 12 Hospital, University of British Columbia, Vancouver, BC, Canada 3 Division of Allergy, Asthma, and Clinical 13 Immunology, Mayo Clinic, Scottsdale, AZ and Division of Pulmonology Phoenix Children’s Hospital 14 Phoenix, AZ; 4Division of Gastroenterology, Department of Medicine and Department of 15 Healthcare Research and Policy. Weill Cornell Medicine; 5Division of Gastroenterology, Alta Bates 16 Summit Medical Center, Oakland, California;; 6Division of Allergy and Immunology, Nationwide 17 Children's Hospital and The Ohio State University College of Medicine, Columbus, Ohio, United States; 18 7Division of Gastroenterology and Hepatology, Mayo Clinic College of Medicine, Rochester, Minnesota 19 8 Section of Allergy/Immunology, Children’s Hospital Colorado, University of Colorado School of 20 Medicine, Aurora, CO; 9 Division of Gastroenterology and Hepatology, Cleveland VA Medical Center 21 and University Hospitals, Case Western Reserve University School of Medicine, Cleveland, OH, 10 22 Department of Internal Medicine, University of Cincinnati College of Medicine, Cincinnati, OH; 11 23 Division of Allergy and Asthma, Stanford University School of Medicine, Palo Alto, CA; 12 Department 24 of Medicine, Johns Hopkins University, Baltimore, MD; 13 Division of Allergy & Immunology, 25 Department of Medicine, University of Texas Southwestern Medical Center, Dallas, TX; 14 Division of 26 Allergy and Immunology, The University of Tennessee Health Science Center, LeBonheur Children’s 27 Hospital, Memphis, TN; 15 Department of Internal Medicine, New Jersey Medical School, Morristown, 28 NJ; 16 Section of Allergy and Immunology, Dartmouth-Hitchcock Medical Center and Dartmouth Geisel 29 School of Medicine, Lebanon and Hanover, NH; 17 Department of Medicine, Nova Southeastern 30 University, Davie, FL; 18 Division of Allergy and Immunology, Department of Pediatrics, The Elliot and 31 Roslyn Jaffe Food Allergy Institute, Icahn School of Medicine at Mount Sinai, Kravis Children’s 32 Hospital, New York NY.

33

34 Acknowledgements: Clinical Guidelines Committee included:

35

36 This document presents the official recommendations of the American Gastroenterological Association

37 (AGA) and the Joint Task Force on Allergy-Immunology Practice Parameters on the Management of

1

38 Eosinophilic Esophagitis. The guideline was jointly developed by the AGA’s Clinical Practice Guideline

39 Committee and the Joint Task Force on Allergy Practice Parameters with approval of both the boards of

40 the American Academy of Allergy, Asthma, and Immunology and the American College of Allergy,

41 Asthma, and Immunology; and approved by both the AGA Governing Board and JTF Governing

42 Boards. Development of this guideline and its accompanying technical review was fully funded by both

43 the AGA Institute and the JTF, with no additional outside funding.

44

45 1. Disclose conflicts of interest:

46 Conflict of interest disclosure: All members were required to complete disclosure statement. By mutual

47 agreement with the JTF ,these statements are maintained at the American Gastroenterological

48 Association Institute (AGA) headquarters in Bethesda, Maryland and pertinent disclosure are published

49 with the report.

50

51 Dr. Hirano is supported by NIH grants U54AI117804 and 1P01DK117824 and has received consulting

52 fees and research support from Adare, Allakos, Celgene, Regeneron, Shire Pharmaceuticals. Dr. Chan

53 has received research support from DBV Technologies, has been a member of advisory boards for Pfizer

54 and Aralez Pharmaceuticals, is a member of the scientific advisory board for Food Allergy Canada, and

55 was an expert panel and coordinating committee member of the National Institute of Allergy and

56 Infectious Diseases (NIAID)-sponsored Guidelines for Peanut Allergy Prevention. Dr. Sharaf receives

57 salary support from NCI 1K07CA216326 - 01A1, NCI 5 R01 CA211723 02, and a PCORI Improving

58 Health Systems Award. He is a paid consultant for the non-profit Institute for Clinical and Economic

59 Review. Dr. Stukus received consulting fees from Aimmune Therapeutics, Inc. and Before Brands to

60 deliver unbranded educational symposia. Dr. Greenhawt is supported by grant #5K08HS024599-02

61 from the Agency for Healthcare Quality and Research, is an expert panel and coordinating committee

2

62 member of the NIAID-sponsored Guidelines for Peanut Allergy Prevention; has served as a consultant

63 for the Canadian Transportation Agency, Thermo Fisher, Intrommune, and Aimmune Therapeutics; is a

64 member of physician/medical advisory boards for Aimmune Therapeutics, DBV Technologies,

65 Sanofi/Genzyme, Genentech, Nutricia, Kaleo Pharmaceutical, Nestle, Aquestive, Allergy Therapeutics,

66 and Monsanto; is a member of the scientific advisory council for the National Peanut Board; has

67 received honorarium for lectures from Thermo Fisher, Aimmune, DBV, Before Brands, multiple state

68 allergy societies, the ACAAI, the EAACI; is an associate editor for the Annals of Allergy, Asthma, and

69 Immunology; and is a member of the Joint Taskforce on Allergy Practice Parameters. These

70 relationships are unrelated to the work on this guideline and pose no conflict of interest. The

71 recommendations involving medications undergoing clinical trials were written by members of the

72 guideline committee without conflict of interest.

73

74

75

3

76 Eosinophilic esophagitis (EoE) was first characterized as a distinct clinical entity by Attwood

77 and Straumann in the early 1990s. A dramatic rise in the recognition of EoE in the United States, first in

78 pediatrics and subsequently in adults, was paralleled by an increase in publications on EoE.1 The past 25

79 years has witnessed the emergence of the field from small case series and observational studies to larger,

80 international, multi-center, randomized controlled trials (RCTs) of both medical and dietary therapies.2

81 This guideline provides evidence-based recommendations focusing on the clinical management of EoE

82 for both pediatric and adult allergists and gastroenterologists. Unless specified, the recommendations are

83 applicable to the short-term treatment of EoE, as the current evidence-base is primarily comprised of

84 trials extending from 2 to 16 weeks. The majority of recommendations are based on the failure to

85 achieve histologic remission of < 15 eosinophils/high power field (eos/hpf) as the definition of treatment

86 effect.2 Additional relevant outcome metrics, including symptoms and endoscopic features, could not be

87 synthesized due to the use of varying and largely unvalidated instruments, variable study methodology,

88 and a large degree of heterogeneity in reporting of outcomes. In forming the estimate of the effect for

89 observational studies lacking a contemporaneous control group, the 8-week, placebo control arm rate for

90 failing to achieve histologic remission from topical glucocorticosteroid studies (86.7%) was used to

91 allow comparison. In recommendations that this historical control group was used, the quality and

92 strength of evidence was downgraded for using this indirect comparator. For these recommendations,

93 risk ratios (RRs) are presented by applying the baseline risk from the untreated control arms from steroid

94 RCTs to the RR. As was reviewed in the technical report use of this comparator should not be viewed

95 the same as a direct control group comparison, but as an approximated measure that is permissible under

96 GRADE methodology.

97

98 The guideline was developed utilizing a process outlined elsewhere.3 Briefly, both the AGA and

99 the JTF process for developing clinical practice guidelines incorporates GRADE methodology3 and best

4

100 practices as outlined by the Institute of Medicine.4 GRADE methodology was utilized to prepare the

101 background information for the guideline and the technical review which accompanies it.2 GRADE uses

102 the PICO format, which frames a clinical question by defining a specific Population (P), Intervention (I),

103 Comparator (C), and Outcomes (O). The PICOs focused on the use of therapeutics in patients with EoE.

104 Each of the selected PICO questions was addressed in this review using the GRADE framework except

105 for last two PICO questions which were addressed using a narrative review format. Optimal

106 understanding of this guideline will be enhanced by reading applicable portions of the technical review.

107 A unique aspect of this guideline and the corresponding technical review was the development in a joint

108 manner through a collaboration between AGA and Joint Task Force for Allergy-Immunology Practice

109 Parameters (JTF) that is comprised of the American Academy of Allergy, Asthma & Immunology

110 (AAAAI), American College of Allergy, Asthma and Immunology (ACAAI). In addition, representatives

111 of both pediatric and adult medicine were included as well as a patient with EoE. This collaborative

112 guideline reflects the interdisciplinary nature of EoE that integrates clinical and investigative efforts of

113 multiple domains and builds upon prior consensus recommendations published in both the allergy and

114 gastroenterology literature.5,6

115

116 117 Table 1. GRADE Definitions on Strength of Recommendation

For the Patient For the Clinician

Strong Most individuals in this situation would want the Most individuals should receive the recommended course of action and only a small recommended course of action Formal decision proportion would not. aids are not likely to help individuals make decisions consistent with their values and preferences. Conditional The majority of individuals in this situation would Different choices will be appropriate for different want the suggested course of action, but many patients. Decision aids may be useful in helping would not. individuals in making decisions consistent with their values and preferences. Clinicians should expect to spend more time with patients when working towards a decision. 118

5

119

120 Table 2. GRADE Definitions on Quality of Evidence

High We are very confident that the true effect lies close to that of the estimate of the effect. Moderate We are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low Our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect. Very We have very little confidence in the effect estimate. The true effect is likely to be substantially different from Low the estimate of effect 121

122 Table 3: AGA-JTF Guideline Recommendations on the Management of Eosinophilic Esophagitis (EoE)

Recommendation Strength of Quality of Evidence Recommendation 1. Recommendation: In patients with symptomatic esophageal eosinophilia, the AGA/JTF suggests using Conditional Very low quality proton pump inhibition over no treatment

2. In patients with EoE, the AGA/JTF recommends topical Strong Moderate glucocorticosteroids over no treatment 3. In patients with EoE, the AGA/JTF suggests topical Conditional Moderate glucocorticosteroids rather than oral glucocorticosteroids 4. In patients with EoE, the AGA/JTF suggests using elemental diet over no treatment

Comment: Patients who put a higher value on avoiding the Conditional Moderate challenges of adherence to an elemental diet and the prolonged process of dietary reintroduction may reasonably decline this treatment option.

5. In patients with EoE, the AGA/JTF suggests using an empiric, six-food elimination diet over no treatment

Comment: Patients who put a higher value on avoiding the Conditional Low challenges of adherence to diet involving elimination of multiple common food staples and the prolonged process of dietary reintroduction may reasonably decline this treatment option. 6. In patients with EoE, the AGA/JTF suggests using an allergy testing-based elimination diet over no treatment

Comment: Due to the potential limited accuracy of currently available, allergy-based testing for the Conditional Very low quality identification of specific food triggers for EoE, patients may prefer alternative medical or dietary therapies to an exclusively testing-based elimination diet.

7. Recommendation: In patient with EoE in remission following short-term use of topical glucocorticosteroids, the Conditional Very low quality AGA/JTF suggests continuation of topical glucocorticosteroids over discontinuation of treatment

6

Comments: Patients who put a high value on the avoidance of long-term topical steroid use and its possible associated adverse effects, and/or place a lower value on the prevention of potential long-term undesirable outcomes (i.e. recurrent dysphagia, food impaction, and esophageal stricture), could reasonably prefer cessation of treatment after initial remission is achieved, provided clinical follow- up is maintained.

8. Recommendation: In adult patients with dysphagia from Conditional Very low quality a stricture associated with EoE, the AGA/JTF suggests endoscopic dilation over no dilation

9. Recommendation: In patients with EoE, the AGA/JTF recommends using anti-IL-5 therapy for EoE only in the context of a clinical trial No recommendation Knowledge gap

10. Recommendation: In patients with EoE, the AGA/JTF recommends using anti-IL-13 or anti-IL-4 receptor alpha therapy for EoE only in the context of a clinical trial No recommendation Knowledge gap

11. Recommendation: In patients with EoE, the AGA/JTF suggests against the use of anti-IgE therapy for EoE Very low quality Conditional

12-15. Recommendation: In patients with EoE the AGA/JTF suggest using montelukast, cromolyn sodium, immunomodulators, and anti-TNF for EoE only in the No recommendation Knowledge gap context of a clinical trial

123

124

125 RECOMMENDATIONS

126 Question 1. Should proton pump inhibitors be used in patients with esophageal eosinophilia?

In patients with symptomatic esophageal eosinophilia, the AGA/JTF suggests using proton pump inhibition over no treatment (conditional recommendation, very low-quality evidence).

127

7

128 Twenty-three observational studies that evaluated the histologic response to proton pump

129 inhibitors (PPI) reported an overall, unweighted histologic response rate of 42%. PPIs failed to induce

130 histologic remission in approximately 2/3 of treated patients, compared with >85% of patients treated

131 with placebo, for a RR of 0.66, (95% CI 0.61-0.72). A high degree of inconsistency makes it difficult to

132 provide a precise estimate of an absolute effect size and raises important concerns regarding variation in

133 the criteria for patient selection, study design, as well as proton pump inhibitor (PPI) duration, dosing

134 and formulation. Furthermore, most studies were non-comparative, single-arm, retrospective studies.

135 Based on these factors, the strength of the recommendation was lowered. Nevertheless, a clinical

136 benefit to the use of PPI monotherapy may be evident for certain patients. It is important to note that a

137 European and an International consensus recommendation have recently removed the PPI trial from the

138 diagnostic criteria of EoE.7,8 Following the exclusion of secondary causes of esophageal eosinophilia,

139 symptomatic esophageal eosinophilia is now viewed as synonymous with EoE. PPIs are positioned as an

140 effective, primary therapeutic option for certain patients with EoE. Based on their long-standing safety

141 profile and ease of administration, patients may prefer to start with this form of therapy prior to trials of

142 glucocorticosteroids or elimination diets. It should be emphasized that direct comparison of the efficacy

143 of PPI and other medical or dietary EoE therapies is limited since, up to this time, most trials in EoE

144 have excluded patients with esophageal eosinophilia that responded to a PPI (formerly denoted as PPI-

145 responsive esophageal eosinophilia or PPIREE).

146

147 Question 2. Should topical glucocorticosteroids be used in patients with EoE?

148

In patients with EoE, the AGA/JTF recommends topical glucocorticosteroids over no treatment (Strong recommendation, moderate quality evidence)

8

149 Eight double blind placebo-controlled studies enrolling 437 patients followed for a mean of eight

150 weeks compared treatment with topical budesonide or topical fluticasone to placebo.2 It is of note that

151 most of these studies required patients first fail a PPI trial or excluded patients with known GERD,

152 which may not reflect routine clinical practice or the most current consensus-driven recommendations.

153 Two of the trials used formulations of topical steroids specifically developed for esophageal delivery

154 (tablet or liquid) whereas the remainder utilized ingested formulations designed for the treatment of

155 asthma. As the result of a review process described in the technical guidelines, a single pooled estimate

156 is presented herein, despite many methodologic differences between these studies including the relative

157 potency and bioavailability of the agents used, method of administration, definition of response, dose,

158 and differences that may occur in pediatric versus adult patients. All such factors may limit

159 generalizability of this recommendation. Topical glucocorticosteroids failed to induce histologic

160 remission in approximately 1/3 of treated patients, compared with >85% of patients treated with

161 placebo, for a RR of 0.39, (95% CI 0.26-0.58).2 The certainty of this estimate is moderate; it was

162 downgraded for inconsistency due to heterogeneity of the studies. In short-term studies of three months

163 or less, there was no increased risk of adverse events in patients treated with steroids compared with

164 placebo (risk ratio of 1 (95% CI 0.85-1.19), although local viral and fungal infections and very limited

165 description of adrenal suppression have been described in certain populations. Longer-term studies

166 prospectively assessing the safety of topical glucocorticosteroid use are ongoing. It is relevant to

167 consider that the same inhaled steroid agents are considered very safe for use in children and adults with

168 asthma and are routinely used in the primary management of this disease. While no medications have

169 been yet approved for treatment of EoE by the Food and Drug Administration, the European Medicines

170 Agency approved a budesonide tablet formulation for EoE in 2018.

171

9

172 Question 3. Should systemic glucocorticosteroids be used in patients with EoE?

173

In patients with EoE, the AGA/JTF suggests topical glucocorticosteroids rather than oral glucocorticosteroids (Conditional recommendation, moderate quality evidence)

174 There has only been a single randomized trial of topical versus systemic glucocorticosteroids in

175 80 children with EoE.2 Prednisone was given at a dose of 1 mg/kg twice a day while fluticasone was

176 given at a dose of 2 puffs four time a day (110 ug/puff for those < 10 years of age and 220 ug/puff for

177 those 11-18 years) for 4 weeks followed by tapered dosing over 8 weeks. The primary endpoint was the

178 histological response which was based on a score consisting of the percentage of basal cell hyperplasia

179 and eosinophil density (eos/hpf). Both groups had similar histological improvement defined as a 1-point

180 drop in this score. However, this score showed statistically greater improvement in the prednisone

181 treated group compared to the fluticasone treated group at 4 weeks. The clinical significance of this

182 difference, however, is unclear given that symptomatic improvement was similar in both groups with

183 72% response rates in the prednisone arm versus 65% in the fluticasone arm. Relapse rates were also

184 similar at 45% in both groups at week 24. Systemic complications were increased at 40% in the

185 prednisone group, including weight gain and Cushingoid appearance compared with a 15% rate of oral

186 Candidiasis in the fluticasone group. Based on the similar effectiveness and well-characterized side

187 effects of systemic glucocorticosteroids, topical glucocorticosteroids are preferred over prednisone for

188 treatment of children with EoE. Similarities in disease pathogenesis and clinical manifestations in

189 children and adults with EoE support the extension of the recommendation to adult populations. The

190 potential benefits of systemic glucocorticosteroids in EoE patients that are refractory to topical

191 glucocorticosteroids are currently unknown.

192

10

193 Question 4. Should an elemental diet be used in patients with EoE?

In patients with EoE, the AGA/JTF suggests using elemental diet over no treatment (conditional recommendation, moderate quality evidence).

Comment: Patients who put a higher value on avoiding the challenges of adherence to an elemental diet and the prolonged process of dietary reintroduction may reasonably decline this treatment option.

194

195 The relevant data on efficacy of elemental diets (amino acid-based formulas) for treatment of

196 EoE are derived from 6 single arm, observational studies without control group comparators, which

197 indicate that very few (6.4%) of these subjects on elemental diet failed to achieve histologic remission

198 (defined as <15 eosinophils/hpf).2 This contrasts with failure to achieve histologic remission in 86.7%

199 of a historical placebo comparison group glucocorticosteroid, resulting in an estimated RR of 0.07 (95%

200 CI 0.05-0.12).2 Adult studies had a lower proportion of participants achieving histologic remission than

201 pediatric studies.2

202 Difficulty adhering to elemental diets for reasons such as taste, nutritional concerns, practical

203 implementation within the context of overall dietary alternatives, and breadth of avoidance in this style

204 of diet, and cost are of concern. Harms include interference with development of oral motor skills in

205 children, social isolation created by dining restrictions, the potential need for gastrostomy tube, costs of

206 elemental formula, and burden of repeated endoscopies during gradual food re-introduction. From a

207 food allergy perspective, there may be some risk of developing de novo IgE-mediated food allergy in

208 previously tolerant patients on elimination diets for EoE, as has been noted in isolated case reports in

209 EoE as well as in atopic dermatitis.9 10 There is insufficient literature beyond a handful of case reports

210 describing such events to determine if such risk exists and further studies are needed to evaluate this

211 concern. Hence, elimination diets of any type should be used in discretion and for as short a period that

11

212 is suitable to treat the underlying EoE. Consultation with a board-certified allergist who is skilled in the

213 management of IgE-mediated food allergy should be a strong consideration.

214 Therefore, although the evidence for efficacy of elemental diets is of moderate quality due to

215 possible large effects, we suggest a conditional recommendation for elemental diet. Clinicians should

216 consider patient age and preferences for alternative medical and dietary management therapeutic options

217 when considering elemental diets.

218

219 Question 5. Should an empiric food elimination diet be used in patients with EoE?

In patient with EoE the AGA/JTF suggests using an empiric six-food elimination diet over no treatment (conditional recommendation, low quality evidence).

Comment: Patients who put a higher value on avoiding the challenges of adherence to diet involving elimination of multiple common food staples and the prolonged process of dietary reintroduction may reasonably decline this treatment option.

220

221 Ten studies reported the effectiveness of an empiric, six-food elimination diet (SFED) with an

222 overall, unweighted histologic response rate of 68%, though these also suffered from the same

223 limitations of elemental diet studies in that all were single arm, observational studies.2 The RR for

224 failure to achieve histologic remission relative to placebo based on historical controls was 0.38 (0.32 to

225 0.43).2 While uniformly beneficial from these observational studies, certainty in the effect estimate was

226 rated down as none of the studies were controlled trials. Although these studies were reported as “six-

227 food” elimination diets, the inclusion of both tree nuts and peanuts as well as finned fish and shellfish

228 could be considered as 8 separate food groups. Furthermore, this approach entails a higher number of

229 actual foods on account of the multiple different types of tree nuts, fin fish and shellfish. In addition, two

12

230 studies eliminated foods to which patients had abnormal skin testing and one also eliminated corn, rice

231 and legumes.

232 Several practical concerns limit the utilization of empiric elimination diets in EoE. Heterogeneity

233 in response rates could reflect selection bias and potential for exclusion of patients with limited

234 adherence to the diet. Incomplete or inconsistent diet reintroduction may reflect the challenges in

235 adherence and activity assessment defining disease relapse (symptoms vs. pathology) during the

236 reintroduction process. As is common to any form of elimination diet, the time, risk, and financial

237 burden of repeated endoscopies are also potential implementation barriers, as is long-term adherence

238 following the identification of specific food trigger(s) in the re-introduction process and the possible

239 development of de-novo IgE-mediated food allergy upon re-introduction.11

240 Several trials were reviewed utilizing empiric elimination diets that limited the number of

241 avoided foods to 1, 2 or 4 given data by Kagalwalla et al suggestive that peanut/tree nut and

242 finfish/shellfish reintroduction after SFED was associated with low rates of disease recurrence, and that

243 not all major allergens needed to be removed initially.2 While this approach potentially reduces the

244 burden of repeated endoscopies during the reintroduction process and may improve lifestyle and

245 adherence, the effectiveness appears to be lower.2 Less invasive procedures such as transnasal

246 endoscopy (which does not require sedation) as well as non-endoscopic, office-based methods to assess

247 disease activity through assessment of surrogate markers are under development and could obviate the

248 need for repeated biopsy sampling during the reintroduction process, thereby increasing the practical

249 application of elimination diet for EoE.12

250

13

251 Question 6. Should allergy-based testing be used for the purpose of identifying food triggers in

252 patients with EoE?

6. In patients with EoE, the AGA/JTF suggests allergy testing-based elimination diet over no treatment (Conditional recommendation, very low quality evidence)

Comment: Due to the potential limited accuracy of currently available, allergy-based testing for the identification of specific food triggers, patients may prefer alternative medical or dietary therapies to an exclusively testing-based elimination diet

253

254 Like elemental and empiric elimination diets, the evidence-base for using allergy-based testing to

255 identify food triggers in patients with EoE is limited to single-armed, observational studies that have

256 non-comparative study designs. Testing based diets involve the scientific rational of identifying a

257 potential immune-mediated mechanism of food allergy involving either food-specific IgE or cell-

258 mediated pathways, as opposed to empiric diets which simply presume importance of common allergens

259 as trigger without identifying their direct role in the pathogenesis of the disease process. Twelve single

260 armed studies reported that 49.2% of subjects on an allergy-testing based elimination diet failed to

261 achieve histologic remission (defined as <15 eosinophils/hpf). The estimated RR for failure to achieve

262 histologic remission relative to placebo based on historical controls was 0.57 (0.33 to 0.73).2 An

263 important limitation in pooling these studies involves the degree of inconsistency due to different testing

264 techniques (e.g., skin prick testing, serum specific IgE testing, patch testing, or combinations of these)

265 used in different studies. A sensitivity analysis failed to show any statistically significant difference

266 between studies that used patch testing and those that did not; however, a sensitivity analysis excluding

267 studies using serum specific IgE was not performed. There may be a potential role for aeroallergen

268 testing and treatment in EoE, which is presently being evaluated. Similar to potential risks for other

269 dietary elimination strategies, there may be challenges with long-term adherence to dietary elimination

14

270 and a potential risk of de-novo IgE-mediated food allergy upon re-introduction.

271

272 Question 7: Should maintenance therapy be recommended in patients with EoE?

Recommendation: In patient with EoE in remission following short-term use of topical glucocorticosteroids, the AGA/JTF suggests continuation of topical glucocorticosteroids over discontinuation of treatment (Conditional recommendation, very low quality evidence)

Comments: Patients who put a high value on the avoidance of long-term topical steroid use and its possible associated adverse effects, and/or place a lower value on the prevention of potential long-term undesirable outcomes (i.e. recurrent dysphagia, food impaction, and esophageal stricture), could reasonably prefer cessation of treatment after initial remission is achieved, provided clinical follow-up is maintained.

273

274 The chronicity and potential for disease progression provide the rationale for maintenance

275 therapy of EoE. Retrospective natural history studies, placebo data from randomized controlled trials,

276 and observational cohort studies support the likely chronic nature of symptoms and histopathology of

277 EoE if it either is untreated, or if treatment is discontinued. Spontaneous disease remission has been

278 reported but is uncommon in either pediatric or adult series, with limited description in the literature.

279 Moreover, the available data in adults, albeit retrospective and subject to certain biases, have

280 demonstrated the potential for long-term progression from inflammation to esophageal strictures in a

281 proportion of EoE patients with untreated disease. 1,19

282 At this time, there are a paucity of studies, and therefore very limited evidence, to define what

283 constitutes effective maintenance therapy in EoE.2 Only one very small trial randomized patients to a

284 year of low dose budesonide (0.25 mg bid) or placebo. While a significant reduction in eosinophil

285 density was noted with active drug compared to placebo, only 36% of patients maintained an eosinophil

15

286 density < 5 eos/hpf at one year, and no dose finding study supported the choice of the 0.25mg BID as

287 appropriate or sufficient versus other amounts. The use of a low maintenance dose of budesonide

288 compared to the induction dose of 1 mg BID likely reduced the efficacy, although development of

289 steroid-tolerance or selection of steroid-refractory patients is plausible. Additional single-armed

290 observational studies of topical steroids also reported a high proportion of patients with histologic

291 recurrence but most also utilized dosing lower than administered during induction. In contrast, three,

292 single-armed observational studies of PPIs noted sustained histologic response in the majority of adults

293 despite dose reduction. Very limited data are available on the long-term effectiveness of elimination

294 diets.

295 Until more data are available, the continued use of either PPI, topical glucocorticosteroids or

296 elimination diets are reasonable options, and this is a very preference-sensitive area of management. As

297 there was limited evidence on PPI or diet therapies, the guideline recommendation was written to only

298 include topical glucocorticosteroids. The limited data, as well as uncertainties in the natural history of

299 EoE, provide very low confidence in the estimated benefits of long-term therapy for EoE, but must also

300 be balanced with the risks of potential disease recurrence in individual patients when treatment is

301 discontinued.

302

303 Question 8. Should esophageal dilation be used in patients with EoE?

304

In adult patients with dysphagia from a stricture associated with eosinophilic esophagitis, the AGA/JTF suggests endoscopic dilation over no dilation (Conditional recommendation, very low-quality evidence.)

305 A recent meta-analysis of 1607 patients in 39 publications who underwent esophageal dilatation

306 found symptom relief in 85% in patients with dysphagia and EoE, although this evidence was

307 considered low quality due to the retrospective, single arm design of these reports and the lack of a

16

308 standard definition for what constitutes clinical improvement.2,13 There is no associated benefit in terms

309 of histological improvement in eosinophilia with dilation, and dilation is considered a point of care

310 option for the endoscopist. Of importance, despite the initial case reports of increased complications

311 from dilation in EoE, large series using a more conservative dilatation approach in experienced centers

312 found that complications were not increased over rates expected from dilation of non-EoE, benign

313 esophageal strictures.2

314 A systematic review and meta-analysis of 977 patients who underwent 2034 dilatations for EoE

315 found that the rate of perforation was 0.033% (95% CI 0-0.226%) with none of the perforations

316 requiring surgical intervention nor related to subsequent patient mortality.2,14 Most of the perforations

317 were prior to 2009 with subsequent improvement in perforation rate after this time period which was

318 speculated to be the result of more conservative techniques. The use of balloon or bougies did not

319 significantly increase rate of perforation though numerically balloons had a perforation rate of 0.06%

320 (95% CI 0 – 0.37%) versus 0.02% (95% CI 0-0.34%) with bougies. Larger dilators (>17 mm diameter)

321 were associated with an increased rate of perforation though this was not statistically significant since

322 only 9 perforations were reported in these studies. The most common adverse event reported was chest

323 discomfort in 24% (95% CI 5.9-41%) and GI hemorrhage occurring in 0.03% (95% CI 0-0.217%).

324 For individual patients that place a higher value on the avoidance of the uncommon

325 complications of dilation, it may be reasonable to use medical or dietary therapy prior to using dilation.

326 Though fibrostenotic disease may be present in many of these patients, it has not been demonstrated that

327 these patients will respond better to dilation as opposed to alternative medical or diet therapies.

328 Retrospective case series have identified a lower utilization of esophageal dilation among patients

329 effectively treated with medical therapy. Esophageal dilatation alone as a treatment modality for patients

330 with EoE and daily dysphagia has only been reported in a small retrospective series and required

331 maintenance dilatation on average every two years.15 The limited, available data support the use of

17

332 medical/diet therapy in combination with periodic dilation as necessary for adults with EoE and

333 esophageal stricture.

334 Question 9. Should anti-IL-5 therapy be used in patients with EoE?

335

In patients with EoE the AGA/JTF recommends using anti-IL-5 therapy only in the context of a clinical trial (no recommendation; knowledge gap).

336 Given the role of IL-5 in the maturation and release of eosinophils, there is a biologically

337 plausible mechanism to support the use of anti-IL-5 therapy in patients with EoE. Three RCTs have

338 been conducted, 2 using mepolizumab (1 involving adults and 1 in children) and 1 using reslizumab

339 (children).2 Participants in each study had higher baseline levels of esophageal eosinophilia and had

340 frequently failed clinical management with other therapies. The results from the mepolizumab and

341 reslizumab studies were combined for the purpose of GRADE analysis despite difference in ages of

342 enrollees of these trials and similar mechanisms of action of these therapies though formal non-

343 inferiority between the drugs has not been studied. While a reduction in tissue eosinophilia was

344 observed, very few participants achieved prespecified histologic remission with < 15 eosinophils/hpf.

345 Greater than 90% of patients in treatment groups failed to achieve histologic remission, for a RR of 0.92

346 (0.84-1.00). Symptomatic improvement was evaluated differently in each study and not grouped for

347 GRADE analysis; however, a significant improvement in symptoms compared with placebo was not

348 observed. No significant safety issues occurred in any of the trials.

349

350 Anti-IL-5 therapies are currently approved for use in moderate-to-severe persistent eosinophilic

351 asthma. Initial studies in asthmatics demonstrated a reduction in tissue eosinophilia but lack of clinical

352 improvement. Follow-up studies that focused treatment on a more specific patient population with

353 steroid resistant refractory eosinophilic asthma were needed to better understand potential clinical

18

354 benefit. In a similar fashion, additional studies in patients with EoE are needed before use in clinical

355 practice can be recommended.

356

357 Question 10. Should anti-IL-13 therapy be used in patients with EoE?

In patients with EoE the AGA/JTF recommends using anti-IL-13 or anti-IL4 receptor alpha therapy for EoE only in the context of a clinical trial (no recommendation; knowledge gap).

358

359 The IL-4 and IL-13 pathway is known to be involved in Th-2 inflammatory conditions by

360 directing eosinophil production, prolonged survival, and trafficking into tissues. Anti-IL-4 and anti-IL-

361 13 therapy has shown benefit in Th-2 associated conditions such as atopic dermatitis and asthma, thus

362 there is a biologically plausible pathway for use in EoE. IL-13 is overexpressed in the esophageal

363 mucosa and induces a gene expression profile that closely resembles the EoE transcriptome.

364 Three clinical trials have evaluated the efficacy of biologic therapy directed at the IL-13 pathway

365 in EoE. One RCT involving 25 adult participants evaluated the use of anti-IL-13 therapy with QAX576

366 in EoE.2 This study did not meet its primary endpoint of > 75% decrease in peak eosinophil counts 12

367 weeks after starting therapy. Mean esophageal eosinophil counts decreased compared with placebo, but

368 no significant difference was observed in symptoms. Two additional RCTs that utilized monoclonal

369 antibodies targeting the IL-13 pathway were not included as the full manuscripts were not available at

370 the time of this systematic review, both of which showed promise. The first was a phase 2 study using

371 RPC4046, a monoclonal antibody against IL-13, that demonstrated histologic and endoscopic efficacy

372 compared to placebo in 99 adults with EoE.16 The second study using dupilumab, a monoclonal

373 antibody against the IL-4α receptor inhibiting the signaling of both IL-13 and IL-4, demonstrated

374 symptom, histologic and endoscopic efficacy compared to placebo in 47 adults with EoE.3 While these

19

375 newer preliminary results appear favorable, the use of anti-IL-13 therapy in EoE is not recommended for

376 clinical use outside of a clinical trial at this time.

377 Question 11. Should anti-IgE therapy be used in patients with EoE?

In patients with EoE the AGA/JTF suggests against the use of anti-IgE therapy for EoE (conditional recommendation; very low-quality evidence)

In patients with EoE, the AGA/JTF recommends topical steroids over no treatment (strong 378 IgE is involved in anti-helminthic responses and mediates type 1 hypersensitivity reactions. recommendation, moderate quality evidence). 379 However, IgE is not known to be directly involved in the development or recruitment of eosinophils.

380 Anti-IgE therapy is currently approved for use in patients with moderate-to-severe persistent atopic

381 asthma and in patients with chronic urticaria who are refractory to first-line therapy.

382 There has been 1 RCT involving 30 adult participants evaluating use of anti-IgE therapy in EoE.2

383 This study did not demonstrate any change in esophageal eosinophilia or reduction in symptoms. Based

384 upon limited evidence and lack of a biologically plausible mechanism, use of anti-IgE therapy in EoE is

385 not recommended for clinical use. A conditional recommendation against use was made for anti-IgE

386 therapy because of the quality of the RCT and the inclusion of the primary endpoint evaluated in this

387 guideline (eosinophils < 15/hpf). Other interventions, such as montelukast (did not include

388 eosinophils/hpf) and cromolyn sodium (very low sample size), had very major limitations and therefore

389 insufficient evidence to recommend against and therefore no recommendation was made about their use.

390 Questions 12-15. Should montelukast, cromolyn, immunomodulators, or anti-TNF therapy be

391 used in patients with EoE?

In patients with EoE the AGA/JTF suggest using montelukast, cromolyn sodium, immunomodulators, and anti-TNF only in the context of a clinical trial (no recommendation; knowledge gap).

20

392

393 Given the few studies and low quality of evidence, use of montelukast, cromolyn,

394 immunomodulators, and anti-TNF therapies are not recommended for clinical use.2 These therapeutic

395 agents have been grouped together for the purposes of this guideline based upon limited evidence for a

396 mechanistic role of their biological markers in the development of EoE and limited studies surrounding

397 each therapy.

398 Montelukast is a leukotriene receptor antagonist approved for use in the treatment of persistent

399 asthma and exercise-induced bronchospasm. There is 1 RCT with adult participants (n=41) comparing

400 montelukast with placebo for maintenance therapy after histologic remission was already achieved and

401 did not show any difference in symptoms. A histologic outcome was not included in the study design.

402 Cromolyn is a mast cell stabilizer that can prevent the release of inflammatory mediators in

403 patients with allergic rhinitis and asthma. Mast cell and mast cell mediators have been implicated in

404 EoE pathogenesis. There has been 1 RCT of cromolyn compared to placebo (n=16), which demonstrated

405 that only 1 patient treated with cromolyn achieved histologic remission.

406 Two immunomodulators (azathioprine and 6-mercaptopurine) have been retrospectively reported

407 in a total of 4 patients with EoE but without any use of control subjects. All patients had EoE refractory

408 to other therapies and multiple confounders that make it difficult to discern the impact of

409 immunomodulatory therapy.2

410 TNF-related apoptosis-inducing ligand has been shown to promote inflammation in EoE. One

411 observational case-series described open-label use of anti-TNF in a clinical trial in 3 adult patients with

412 EoE, all of whom had inadequate response to prior therapy.2 The 3 case reports all reported different

413 outcomes including symptom scores, esophageal eosinophilia and endoscopic changes. While interval

21

414 improvement was observed, the differences in patient presentation, outcome measures and lack of

415 control subjects limit extrapolation of these findings.

416

417 Question 16. Should repeat EGD be used to assess patients with EoE after a change in treatment?

418

419 Numerous randomized, placebo-controlled trials of medical therapies for EoE included in this

420 guideline and accompanying technical review have demonstrated significant improvement in symptom,

421 endoscopic, and histologic endpoints using validated instruments.2 Generally, the improvement in

422 objective parameters of endoscopy and pathology have been more robust and consistent than the

423 subjective improvement in symptom outcomes. Moreover, symptom and pathology outcomes are often

424 discordant with one another, although disease remission currently remains anchored in histologic

425 criteria. Evidence that the assessment of biologic activity with endoscopic and histologic parameters

426 following treatment will reduce long-term disease complications is limited.17 On the other hand, the use

427 of symptom-based therapeutic assessment without EGD and biopsy is limited and often misleading due

428 to the ability of patients to modify dietary intake (i.e. avoidance of hard texture foods, excessive

429 mastication, prolonged meal times) to overcome objective histologic and endoscopic disease

430 manifestations.2,18 Dissociation between biologic activity and symptoms in adults is further compounded

431 by the presence of strictures related to fibrostenosis that do not reflect mucosal inflammatory activity.

432 This concept is evident in the symptom relief provided by esophageal dilation in the absence of

433 improvement in esophageal inflammation.

434 The importance of the documentation of adequate suppression of mucosal inflammation after

435 therapeutic intervention is indirectly supported by several retrospective studies that have associated

436 prolonged, untreated disease with the increased prevalence of esophageal strictures.1 In addition,

22

437 retrospective case series have reported a reduction in frequency of esophageal dilation as well as food

438 impactions with improvement in pathology with topical glucocorticosteroids. Nevertheless, the

439 supposition that reduction in esophageal eosinophilia will prevent progressive disease remodeling

440 consequences requires confirmation in prospective, long-term studies. Similarly, although the use of

441 endoscopic outcomes in GERD and inflammatory bowel disease serve as precedents, their application to

442 EoE needs further study to demonstrate their relevance to long-term disease outcomes.

443 While not a formal recommendation of this guideline, the use of repeat EGD with biopsy to

444 assess disease activity after a change in therapy is reasonable. The criteria for histologic and endoscopic

445 improvement following therapy are being actively investigated to identify core outcome metrics for both

446 clinical trials and clinical practice.2 Until such metrics are established, a threshold of < 15 eos/hpf to

447 define an adequate therapeutic response serves as a response criterion until a better measure is

448 established.3,19 The recommended frequency for EGD with biopsy during clinical follow-up is identified

449 as a knowledge gap and may vary depending upon the severity of initial clinical presentation.

450

451 Question 17. What is the management of patients who become asymptomatic after initial PPI

452 treatment?

453 The recently published European and International consensus statements have removed the PPI

454 trial from the diagnostic criteria for EoE.7,8 Based on this revised definition of EoE, the use of repeat

455 EGD with biopsy after PPI therapy would follow the same rationale as recommendation 16.

456

457 Conclusions

23

458 Over the past 2 decades, EoE has emerged as a dominant cause of dysphagia worldwide. In

459 concert with the rise in disease prevalence, an increasingly robust evidence base has provided insights

460 into effective management strategies that are summarized in this guideline. While swallowed, topical

461 glucocorticosteroids were the only therapy to receive a strong recommendation, the evidence supported

462 conditional recommendations for proton pump inhibition and diet therapy as well as esophageal dilation.

463 The use of novel, targeted biologic therapies for EoE are being actively evaluated. A common theme

464 apparent in both the guideline and the accompanying technical review includes the need for uniform

465 endpoints in clinical trials to facilitate meaningful comparisons between therapies. Furthermore, a

466 deeper understanding of the natural history of EoE in both children and adults is needed to inform

467 clinical decisions regarding the optimal use of disease monitoring and long-term, maintenance therapy.

468 In the dawn of this new disease, much light has been shed and the future is bright.

469

470

471

24

472 REFERENCES

473 1. Dellon ES, Hirano I. Epidemiology and Natural History of Eosinophilic Esophagitis. Gastroenterology. 474 2018;154(2):319-332 e313. 475 2. Rank MA, Sharaf RN, Furuta GT, et al. Technical Review on the Management of Eosinophilic Esophagitis: 476 A report from the American Gastroenterological Association Institute and the Joint Task Force on 477 Allergy-Immunology Practice Parameters Gastroenterology. 2018. 478 3. Sultan S, Falck-Ytter Y, Inadomi JM. The AGA institute process for developing clinical practice guidelines 479 part one: grading the evidence. Clin Gastroenterol Hepatol. 2013;11(4):329-332. 480 4. Clinical Practice Guidelines We Can Trust. National Academies Press (US); 2011. 481 5. Furuta GT, Liacouras CA, Collins MH, et al. Eosinophilic esophagitis in children and adults: a systematic 482 review and consensus recommendations for diagnosis and treatment. Gastroenterology. 483 2007;133(4):1342-1363. 484 6. Liacouras CA, Furuta GT, Hirano I, et al. Eosinophilic esophagitis: updated consensus recommendations 485 for children and adults. J Allergy Clin Immunol. 2011;128(1):3-20 e26; quiz 21-22. 486 7. Molina-Infante J, Bredenoord AJ, Cheng E, et al. Proton pump inhibitor-responsive oesophageal 487 eosinophilia: an entity challenging current diagnostic criteria for eosinophilic oesophagitis. Gut. 488 2016;65(3):524-531. 489 8. Dellon ES, Liacouras CA, Molina-Infante J, et al. Updated International Consensus Diagnostic Criteria for 490 Eosinophilic Esophagitis: Proceedings of the AGREE Conference. Gastroenterology. 2018;155(4):1022- 491 1033 e1010. 492 9. Chang A, Robison R, Cai M, Singh AM. Natural History of Food-Triggered Atopic Dermatitis and 493 Development of Immediate Reactions in Children. J Allergy Clin Immunol Pract. 2016;4(2):229-236 e221. 494 10. Ho HE, Chehade M. Development of IgE-mediated immediate hypersensitivity to a previously tolerated 495 food following its avoidance for eosinophilic gastrointestinal diseases. J Allergy Clin Immunol Pract. 496 2018;6(2):649-650. 497 11. Wang R, Hirano I, Doerfler B, Zalewski A, Gonsalves N, Taft T. Assessing Adherence and Barriers to Long- 498 Term Elimination Diet Therapy in Adults with Eosinophilic Esophagitis. Dig Dis Sci. 2018;63(7):1756-1762. 499 12. Hiremath G, Gupta SK. Promising Modalities to Identify and Monitor Eosinophilic Esophagitis. Clin 500 Gastroenterol Hepatol. 2017;15(11):1655-1664. 501 13. Moole H, Jacob K, Duvvuri A, et al. Role of endoscopic esophageal dilation in managing eosinophilic 502 esophagitis: A systematic review and meta-analysis. Medicine (Baltimore). 2017;96(14):e5877. 503 14. Dougherty M, Runge TM, Eluri S, Dellon ES. Esophageal dilation with either bougie or balloon technique 504 as a treatment for eosinophilic esophagitis: a systematic review and meta-analysis. Gastrointest Endosc. 505 2017;86(4):581-591 e583. 506 15. Lipka S, Keshishian J, Boyce HW, Estores D, Richter JE. The natural history of steroid-naive eosinophilic 507 esophagitis in adults treated with endoscopic dilation and proton pump inhibitor therapy over a mean 508 duration of nearly 14 years. Gastrointest Endosc. 2014;80(4):592-598. 509 16. Hirano I, Collins MH, Assouline-Dayan Y, et al. RPC4046, a Monoclonal Antibody Against IL13, Reduces 510 Histologic and Endoscopic Activity in Patients With Eosinophilic Esophagitis. Gastroenterology. 2018. 511 17. Hirano I, Aceves SS. Clinical implications and pathogenesis of esophageal remodeling in eosinophilic 512 esophagitis. Gastroenterol Clin North Am. 2014;43(2):297-316. 513 18. Safroneeva E, Straumann A, Coslovsky M, et al. Symptoms Have Modest Accuracy in Detecting 514 Endoscopic and Histologic Remission in Adults With Eosinophilic Esophagitis. Gastroenterology. 515 2016;150(3):581-590 e584.

25

516 19. Reed CC, Wolf WA, Cotton CC, et al. Optimal Histologic Cutpoints for Treatment Response in Patients 517 With Eosinophilic Esophagitis: Analysis of Data From a Prospective Cohort Study. Clin Gastroenterol 518 Hepatol. 2018;16(2):226-233 e222.

519

26