aga institute - naspghan

A

ECSG

GPo

*‡

PHUMMGU

Diw(sotEtcgodmlcnacolomdfrpps

Ed

AG

AIN

STITUTE

GASTROENTEROLOGY 2007;133:1342–1363

GA INSTITUTE

osinophilic Esophagitis in Children and Adults: A Systematic Review andonsensus Recommendations for Diagnosis and Treatment

ponsored by the American Gastroenterological Association (AGA) Institute and North American Society of Pediatricastroenterology, Hepatology, and Nutrition

LENN T. FURUTA,* CHRIS A. LIACOURAS,‡ MARGARET H. COLLINS,§ SANDEEP K. GUPTA,� CHRIS JUSTINICH,¶

HIL E. PUTNAM,# PETER BONIS,** ERIC HASSALL,‡‡ ALEX STRAUMANN,§§ MARC E. ROTHENBERG,� � and Membersf the First International Gastrointestinal Eosinophil Research Symposium (FIGERS) Subcommittees##

Section of Pediatric Gastroenterology, Hepatology, and Nutrition, The Children’s Hospital, Denver, University of Colorado Medical School, Denver, Colorado;Division of Gastroenterology, Hepatology, and Nutrition, University of Pennsylvania School of Medicine, The Children’s Hospital of Philadelphia, Philadelphia,ennsylvania; §Division of Pathology, University of Cincinnati, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio; �Division of Pediatric Gastroenterology,epatology, and Nutrition, Riley Hospital for Children, Indiana University School of Medicine, Indianapolis, Indiana; ¶Division of Pediatric Gastroenterology, Queen’sniversity, Kingston General Hospital, Kingston, Ontario, Canada; #Division of Pediatric Gastroenterology, Hepatology, and Nutrition, Cincinnati Children’s Hospitaledical Center, University of Cincinnati College of Medicine, Cincinnati, Ohio; **Division of Gastroenterology, Tufts University School of Medicine, Boston,assachusetts; ‡‡Division of Gastroenterology, BC Children’s Hospital, University of British Columbia, Vancouver, British Columbia, Canada; §§Department of

� �
astroenterology, University Hospital Basel, University Basel, Basel, Switzerland; Division of Allergy and Immunology, Cincinnati Children’s Hospital Medical Center,niversity of Cincinnati College of Medicine, Cincinnati, Ohio; and ##FIGERS Subcommittee listed in Appendix 1
dwctddp

appiFS2batcti

a

tpi

uring the last decade, clinical practice saw a rapidncrease of patients with esophageal eosinophilia whoere thought to have gastroesophageal reflux disease

GERD) but who did not respond to medical and/orurgical GERD management. Subsequent studies dem-nstrated that these patients had a “new” diseaseermed eosinophilic esophagitis (EE). As recognition ofE grew, so did confusion surrounding diagnostic cri-

eria and treatment. To address these issues, a multidis-iplinary task force of 31 physicians assembled with theoal of determining diagnostic criteria and making rec-mmendations for evaluation and treatment of chil-ren and adults with suspected EE. Consensus recom-endations were based upon a systematic review of the

iterature and expert opinion. EE is a clinicopathologi-al disease characterized by (1) Symptoms including butot restricted to food impaction and dysphagia indults, and feeding intolerance and GERD symptoms inhildren; (2) > 15 eosinophils/HPF; (3) Exclusion ofther disorders associated with similar clinical, histo-

ogical, or endoscopic features, especially GERD. (Usef high dose proton pump inhibitor treatment or nor-al pH monitoring). Appropriate treatments include

ietary approaches based upon eliminating exposure toood allergens, or topical corticosteroids. Since EE is aelatively new disease, the intent of this report is torovide current recommendations for care of affectedatients and defining gaps in knowledge for future re-earch studies.

osinophilic esophagitis (EE) is a disease of the esoph-agus that has become increasingly recognized in chil-

ren and adults over the last decade. It is a clinicopathologic

isorder characterized by a dense esophageal eosinophiliaith severe squamous epithelial hyperplasia generally oc-

urring in association with upper gastrointestinal symp-oms, primarily esophageal. In EE, the gastric and duo-enal mucosae are normal. The esophageal abnormalitieso not respond to treatment with high-dose protonump inhibitor (PPI) therapy.Although an increasing number of children and adults

re presenting with EE, few controlled trials have beenerformed to guide management. As a result, clinicalractice is largely based on limited data and expert opin-

on. This review was conducted in preparation for Theirst International Gastrointestinal Eosinophil Researchymposium held in Orlando, FL, on October 17 and 18,006. The clinical recommendations made herein areased on a systematic review of the published literaturend on expert opinion in which there are gaps or con-roversy. The purpose of this review is to document theurrent state of knowledge in EE and to determine howo advance the field by expanding knowledge and defin-ng priorities and strategies for future research.

DefinitionA number of names and acronyms have been

pplied to this disease, including the following: eosino-

Abbreviations used in this paper: EDN, eosinophil-derived neuro-oxin; EE, eosinophilic esophagitis; eos/HPF, eosinophils per high-ower field; GERD, gastroesophageal reflux disease; PPI, proton pump

nhibitors.© 2007 by the AGA Institute and North American Society of Pediatric

Gastroenterology, Hepatology, and Nutrition0016-5085/07/$32.00

doi:10.1053/j.gastro.2007.08.017

peapDsgpfwtAoioc

fwcCsTtFa(emrwmGrt

pcnmloit

tsrlettdwatsbt

bsccrqchrtbaosbp

dcna

s7afpcato

T

GEECCHID

T

C�

L

AG

AIN

STIT

UTE

October 2007 CONSENSUS REPORT OF EOSINOPHILIC ESOPHAGITIS 1343

hilic esophagitis (EE and EoE), primary eosinophilicsophagitis (PEE), allergic eosinophilic esophagitis (AEE),nd idiopathic eosinophilic esophagitis (IEE). For theurposes of this review, we will use the acronym EE.efining EE presents some problems because the pre-

enting symptoms are similar to those of gastroesopha-eal reflux disease (GERD) and include heartburn, chestain, feeding intolerance, dysphagia, odynophagia, andood impaction. However, although GERD may coexistith EE, the symptoms and pathologic features intrinsic

o EE do not respond to acid suppression treatment.lthough basal cell hyperplasia of esophageal mucosaften occurs in EE, as it does in GERD, the distinguish-

ng primary histologic feature of EE is a striking eosin-philia of esophageal mucosa, often with eosinophil mi-roabcesses.

However, esophageal eosinophilia is not exclusivelyound in EE. Among other diseases that are associatedith esophageal eosinophilia are GERD, Crohn’s disease,

ollagen vascular disease, infectious esophagitis (herpes,andida), drug-associated esophagitis, hypereosinophilicyndrome, and eosinophilic gastroenteritis1 (see Table 1).herefore, careful consideration was given to excluding

hese conditions as diagnostic possibilities in the review.or example, one case series described 2 children and 1dult with clinicopathologic features consistent with EEgross evidence of furrowing, white plaques, and �20osinophils per high-power field [eos/HPF] in the squa-ous mucosa) in whom symptoms and histopathology

esolved with intensive PPI medication.2 This is a some-hat unusual example, but, given that it is not uncom-on for patients with EE to have some symptoms ofERD that respond to acid suppression treatment, our

eview focuses on those articles in which EE was clearlyhe primary diagnosis.

For the purposes of this review, EE is defined as arimary clinicopathologic disorder of the esophagus,haracterized by esophageal and/or upper gastrointesti-al (GI) tract symptoms in association with esophagealucosal biopsy specimens containing �15 intraepithe-

ial eos/HPF in 1 or more biopsy specimens and absencef pathologic GERD as evidenced by a normal pH mon-

toring study of the distal esophagus or lack of response

able 1. Differential Diagnosis of Esophageal Eosinophilia

astroesophageal reflux diseaseosinophilic esophagitisosinophilic gastroenteritisrohn’s diseaseonnective tissue diseaseypereosinophilic syndrome

nfectionrug hypersensitivity response

o high-dose PPI medication (see Table 2).N

Methodology of ReviewA task force of 31 physicians who participated in

he First International Gastrointestinal Eosinophil Re-earch Symposium (FIGERS) performed this review. Theeviewers were divided into subcommittees along theines of their recognized expertise in clinical evaluation,ndoscopy, histopathology, allergy, and treatment. A sys-ematic review of the English language medical literaturehrough September 2006 was performed using electronicatabases (MEDLINE, PubMed, and Ovid), with the keyords “eosinophilic esophagitis,” “allergic esophagitis,”nd “eosinophilic oesophagitis.” Review articles, letters tohe editor, most case reports of �3 patients, and ab-tracts were excluded. Several relevant articles on EE haveeen published since the Symposium, and a summary ofhese is provided at the end of this review.

Relevant data were discussed among committee mem-ers in conference calls. Critical evaluations includedtudy design, numbers of patients, definitions used, out-omes reported, and potential biases. The chair of eachommittee synthesized the data, and inconsistencies wereesolved by discussion until consensus was achieved. Theuality of evidence supporting the recommendationsontained in this review was assessed as follows: grade A:omogeneous evidence from multiple, well-designed,andomized (therapeutic) or cohort (descriptive) con-rolled trials, each involving a number of participants toe of sufficient statistical power; grade B: evidence fromt least 1, large, well-designed clinical trial with or with-ut randomization from cohort or case-control analytictudies or well-designed meta-analysis; grade C: evidenceased on clinical experience, descriptive studies, or re-orts of expert committees.The committees determined that the quality of evi-

ence in these articles fell primarily into the grade Category. This finding speaks to the relative recent recog-ition of EE and therefore the need for current guidelinesnd future well-designed, randomized studies.

A total of 80 studies met the inclusion criteria anderve as the basis of this report. They include a total of54 children (age range, 4 months to 20 years) and 323dults (age range, 22– 89 years). The sample sizes variedrom 7 to 381 patients (mean, 37 years). The studies wereublished between 1977 and September 2006. Most wereonducted in academic centers in the United States, Can-da, Europe, and Australia. The following is a review ofhe literature, with critical comments and consensus rec-mmendations on selected topics.

able 2. Diagnostic Guidelines

linical symptoms of esophageal dysfunction15 Eosinophils in 1 high-power field

ack of responsiveness to high-dose proton pump inhibition (up to2 mg/kg/day) or

ormal pH monitoring of the distal esophagus

T31py

treat

eLieae

NlotycaemSiios

dsvgetvp

dcfcost

5maafct11

tftfmr1aaf

yfGonr

iyastrirteCwwd5hppplrpl

AG

AIN

STITUTE

1344 FURUTA ET AL GASTROENTEROLOGY Vol. 133, No. 4

EpidemiologyMales are more commonly affected than females.

hirteen studies provided detailed information regarding23 adult patients (76% males; mean age, 38 years; range,4 – 89 years).3–15 Sixteen studies identified 754 pediatricatients (66% males; mean age, 8.6 years; range, 0.5–21.1ears).16 –31

Geographically, patients with EE have now been iden-ified throughout the United States and Canada, andeports in the literature have originated on all continentsxcept Africa. Although there is frequent discussionbout geographic variations of prevalence, at this time,here are no controlled data to support this.

EE has been described in patients with a variety ofthnic backgrounds, including white, African-American,atin, and Asian, but few studies provided details. Thus,

t remains unclear whether EE is associated with anthnic or racial predilection. Socioeconomic distributionnd seasonal variation in EE have not been systematicallyxamined.

Two studies addressed the increasing prevalence of EE.oel et al identified a 4-fold increase in disease preva-

ence in children with EE in the Midwest United Statesccurring over a period from 2000 to 2003.32 In addition,hey reported an incidence of �1:10,000 children perear, and this incidence remained constant over theourse of the 4-year study. Given the lack of mortalityssociated with EE, the prevalence over time will increaseven if the incidence remains the same. Similarly, Strau-ann and Simon found an increase in EE prevalence in

witzerland from 2 per 100,000 to 27 per 100,000 inhab-tants over a 16-year period.33 It is unlikely that thisncrease can be entirely accounted for by increased rec-gnition because the areas examined were geographicallytable and recording practices were consistent.

A number of reports suggest familial clustering of theisease, but it is difficult to determine whether this ob-ervation represents genetic predisposition or similar en-ironmental exposure.33–35 One study showed that theene encoding the eosinophil-specific chemoattractantotaxin-3 was the most highly induced gene in EE pa-ients compared with its expression level in healthy indi-iduals and may be an indicator of a potential geneticredisposition to EE.29

Clinical ManifestationsChildrenAs with many other diseases, some age-related

ifferences were noted between presenting symptoms inhildren and adults.3,5–23,25–30,36 For instance, feeding re-usal or intolerance is a common symptom of EE inhildren who are perhaps too young to relate the feelingf dysphagia. Children most commonly had GERD-likeymptoms (including heartburn and regurgitation), al-
hough estimates varied widely across studies (range, a
%– 82%). Emesis and abdominal pain were also com-only reported (range, 8%–100% for emesis; 5%– 68% for

bdominal pain). Dysphagia and food impaction werelso reported (range, 16%–100% for dysphagia; 10%–50%or food impaction). These symptoms tended to be in-reasingly common with age. Other presenting symp-oms in children included failure to thrive (range, 5%–9%), chest pain (range, 17%–20%), and diarrhea (range,%–24%).

AdultsIn adults, the most common presenting symp-

oms were intermittent dysphagia (range, 29%–100%) andood impaction (range, 25%–100%). One report foundhat EE was responsible for 50% of cases of esophagealood impaction in one institution.10 Although less com-

on than in children, GERD-like symptoms were alsoeported (range, 7%–100%) as were chest pain (range,%–58%) and abdominal pain (range, 3%–25%). Diarrheand weight loss were reported in some patients. Manydults had long-standing symptoms including recurrentood impactions prior to the diagnosis of EE.10,37

Recommendations. EE should be considered inoung children with GERD-like symptoms, includingeeding problems, and in older children and adults withERD-like symptoms, especially in those with dysphagiar esophageal food impaction. When the primary diag-osis is EE, symptoms are unresponsive or only partiallyesponsive to acid blockade (Grade B).

Natural HistoryThree studies examined the natural history of EE

n 90 adults, with follow-up ranging from 1 to 11.5ears.4,7,11 Potter et al followed 29 patients (21 men; meange, 35 years; range, 16 –71 years) who primarily pre-ented with dysphagia and “refractory GERD” symp-oms.7 The majority of patients showed evidence of tissueemodeling at endoscopy. Rings, strictures, or small cal-ber esophagus were found in 86% of patients, whereasadiographic studies showed narrowing in 67%. Impor-antly, the small caliber esophagus that was observedndoscopically was missed radiographically in 4 patients.roese et al reported their experience with 31 patientsith EE (24 men; mean age, 34 years; range, 14 –77 years)ho most commonly presented with bolus impaction orysphagia (Table 3).11 Diagnosis was delayed a mean of4 months (range, 0 –180 months), and, in retrospect,ighly suggestive features of EE were not recognized in 7atients, leading to a delay in diagnosis. Strictures wereresent in 57% and were described as localized to theroximal esophagus, measuring several centimeters in

ength and extending in a longitudinal fashion. Dilationesulted in longitudinal tears in 77% of patients. Noatient had a perforation. Straumann et al describe the

ongest follow-up of 30 adults with EE (22 men; mean
ge, 40.6 years; range, 16 –71 years).4 The presenting

si4odcpad

oMac6ptiwsdmeewssn

Caeio

Echwmlt

cEut

eerptosetfiituwb

mrwntFsofuEsGm

ptssei

T

C

FV“

“

F

EDF

T

LWCL

S

N

AG

AIN

STIT

UTE


ymptom was almost exclusively dysphagia with foodmpaction, and the diagnosis was delayed an average of.6 years (range, 0 –17 years). During the follow-up periodf 1.4 –11.5 years, 23% of patients reported increasingysphagia, and 36.7% reported stable symptoms. Nohange in endoscopic features was identified in 6 of 7atients in whom a subepithelial component could benalyzed, but an increase in fibrosis and thickening wasocumented.Liacouras et al reported the largest longitudinal study

f 381 children with EE (66% male; mean age, 9 years).22

ost presented with symptoms of GERD refractory tocid suppression treatment or with dysphagia. Upper GIontrast studies demonstrated esophageal narrowing in% of children. Endoscopy showed rings in 12%, and 1atient required esophageal dilation. In a subset of pa-ients, medical treatment with systemic corticosteroidsnduced clinicopathologic remission in all but 1 patient,hereas 52% of patients treated with topical fluticasone

howed improvement, with 2 developing esophageal can-idiasis. Following discontinuation of medical treat-ent, almost all patients had recurring symptoms and

sophageal eosinophilia. Dietary treatment in the form ofither dietary restriction or amino acid-based formulaas highly effective (97.6% showed clinicopathologic re-

ponse) in inducing and maintaining remission. Bariumtudies normalized in 21 of 22 patients with esophagealarrowing.Eosinophilic inflammation seems to persist over time.

linical experience dictates that some patients may besymptomatic and have esophageal eosinophilia. No dataxist as to the best management of these patients, but its suggested that they be followed closely for the devel-pment of clinical symptoms.

The disease does not appear to limit life expectancy.sophageal metaplasia, (that is Barrett’s esophagus orardia-type metaplasia or esophageal adenocarcinoma38)as not been reported in patients with EE, even in adultsith severe disease. EE is not a disease characterized byucosal ulceration or destruction. Therefore, it seems

ikely that the pathologic process of EE is different fromhat of GERD and that adenocarcinoma or squamous

able 3. Symptoms Suggestive of Eosinophilic Esophagitis

hildren Adult

eeding aversion/intolerance Dysphagiaomiting/regurgitation Food impactionGERD refractory to medicalmanagement”

“GERD refractory to medicalmanagement”

GERD refractory to surgicalmanagement”

ood impaction/foreign bodyimpaction

pigastric abdominal painysphagia

Eailure to thrive

ancer of the esophagus are not part of the spectrum ofE, other than perhaps as coincidental occurrences. Nat-ral history and basic studies will provide insights intohe validity of this speculation.

Recommendations. EE tends to be a chronic dis-ase with persistent or relapsing symptoms. To date,sophageal strictures and small caliber esophagus, oftenesulting in food impaction, have been the major com-lications identified. When these findings are encoun-ered, either radiologically or at endoscopy, a high indexf suspicion should be raised for EE, and mucosal biopsypecimens should be obtained (Grade B). Althoughsophageal metaplasia (Barrett’s esophagus or cardia-ype metaplasia) has not been described as an associatednding in patients with EE, careful long-term follow-up

s advised. Other chronic problems include failure tohrive and feeding intolerance in children. At present, it isnclear whether persistent esophageal eosinophilia is al-ays accompanied by symptoms. See Monitoring sectionelow.

Diagnostic TestingEndoscopyAt endoscopy, a number of gross mucosal abnor-

alities have been identified including longitudinal fur-owing, friability, edema, longitudinal shearing, raisedhite specks, whitish exudates, “crêpe paper mucosa,”arrow caliber esophagus, Schatzki ring, felinization, andransient or fixed rings3,6,7,9 –12,22,36,39 – 49 (See Table 4 andigures 1 and 2). All listed findings except longitudinalhearing and “crêpe paper” mucosa have been reported inther esophageal diseases. An earlier report describedeatures such as circular rings that were primarily attrib-ted to GERD but in retrospect likely were related toE.50 This report further emphasizes that a high index ofuspicion for EE must be maintained in any patient withERD-like symptoms who has an abnormal appearingucosa as described above.Although none of the features can be classified as

athognomonic for EE, in the appropriate clinical con-ext, the presence of more than 1 of these findings istrongly suggestive of the diagnosis of EE. In contrast,ome studies have reported normal appearing mucosa. Asxperience has accumulated, particularly in centers focus-ng on endoscopic analysis of children and adults with

able 4. Endoscopic Features Associated With EE

inear furrowing, vertical lines of the esophageal mucosahite exudates, white specks, nodules, granularityircular rings, transient or fixed, felinizationinear shearing/crepe paper mucosa with passage of endoscope ordilator

tricture: proximal, middle, or distal

OTE. None of the features are pathognomonic of EE.

E, subtle abnormalities are now being detected that may

hmo

otaiepsmncshifoii

ddtEeswtpsc

pEastoafooowmfee

idwstm

p

Ftstabau

AG

AIN

STITUTE


ave been previously overlooked in earlier series. Thisay be related to different endoscopic techniques or

lder equipment.

Biopsy Procurement and EvaluationMany earlier studies reported on endoscopic bi-

psy specimens that were taken primarily or solely fromhe distal esophagus. However, over the last several years,n increasing number of studies have included the find-ngs on biopsy specimens from the middle and uppersophagus.6–8,10,11,14,16–18,20,22–25,28,29,32,36,45,48,51–57 Three im-ortant points emerge from these studies. First, severaltudies show that histopathologic abnormalities are com-

on in biopsy specimens obtained from endoscopicallyormal appearing mucosa. For example, in a study of 381hildren with EE, 30% had a normal appearance endo-copically.22 However, over the last decade, as literatureas developed and endoscopists have become increas-

ngly aware of EE, the subtle features have become moreully recognized. Second, Bouin’s preservative was used inne study, and its use resulted in reduced ability to

dentify eosinophils.7 Therefore, fixing mucosal samplesn preservatives other than Bouin’s is preferable. Third, to

igure 1. Endoscopic findings associated with eosinophilic esophagi-is. (A) Mucosal rings representative of transient contractions or fixedtructures. This appearance has also been termed feline esophagus,rachealization, or concentric rings. (B) Whitish exudates scatteredcross the mucosal surface. These represent eosinophilic purulenceurgeoning through the esophageal epithelium. Exudates can appears punctate white nodules, dispersant flocculant material, or in a gran-lar pattern and can occur along the length of the esophagus.

etermine how the number of biopsy samples impactediagnostic ability, Gonsalves et al performed a retrospec-ive analysis of 341 biopsy specimens from 66 adults withE. The results showed that, with a threshold of 15os/HPF, the procurement of 1 biopsy specimen had aensitivity of 55%, in contrast to a sensitivity of 100%ith 5 biopsy specimens.57 Clinical experience suggests

hat areas of gross endoscopic abnormalities, as well asroximal and distal esophageal mucosa (even if macro-copically unremarkable), should be assessed histologi-ally.

Recommendations. The preceding discussionrovides the rationale for histological assessment ofE. Endoscopic appearances should be documentednd photographed. Mucosal pinch biopsy specimenshould be obtained from all patients in whom EE is inhe differential diagnosis. Biopsy specimens should bebtained regardless of the gross appearance of the mucosa,nd multiple biopsy specimens should be obtainedrom different esophageal locations along the lengthf the esophagus. Biopsy specimens should also bebtained from stomach and duodenum to rule outther diseases such as eosinophilic gastroenteritis and,hen appropriate, inflammatory bowel disease. Opti-al fixation is accomplished by using fixative such as

ormalin or paraformaldehyde (Grade C). The cost-ffectiveness of these recommendations has not beenvaluated but deserves further study.

Intraesophageal pH Testing

Data regarding pH monitoring were reportedn 9 studies involving adults and 11 involving chil-ren.3– 6,9,11–20,23,25,27,28,57 Of 228 adults, pH monitoringas performed in 91 (40%) patients, with normal re-

ults in 75 (82%) patients. Of 223 children, pH moni-oring was performed in 173 (78%) patients, with nor-

al results in 156 (90%) patients.

Esophageal Impedance

No impedance monitoring studies have been re-orted in patients with EE.

Figure 2. Esophageal furrowing rep-resentative of mucosal edema andthickening. (A) Furrows encompassthe entire luminal surface of the distalesophagus with a very thick and al-most nodular appearance. (B) In thisFigure, the most prominent furrowingoccurs preferentially along the left lat-eral wall. (C) Vertical lines coursealong the length of the esophagealmucosa. This finding is often mostprominent when the esophagus is in-
flated.

1o7sptsegaa

stllm

GbecAhoEv

praddfSEfsrfimWaatmc

aa

svthaoitpit

tpta

imHhiAemiptf

tir

Ftepba

AG

AIN

STIT

UTE


Esophageal ManometryEsophageal manometry results were reported in

0 studies (7 adult and 3 pediatric) and were performedn 77 adults and 14 children.3,6,9,11,13–15,20,27,28,57 Of the7 adults, the lower esophageal sphincter was normoten-ive in 66, hypotensive in 10, and hypertensive in 1atient. Peristaltic abnormalities were reported in 30 ofhe 77 patients, with 28 of the 30 patients having non-pecific peristaltic abnormalities, and 1 each having distalsophageal spasms and nutcracker esophagus. Esopha-eal manometry, overall, was abnormal in 41 of the 77dult patients (53%). All 14 children had a normal esoph-geal manometry.

Endoscopic UltrasoundEndoscopic ultrasound was performed in one

tudy of 11 children. The study reported significanthickening of the esophageal wall and individual tissueayers, including the combined mucosa and submucosalayer, and the muscularis propria, as compared with nor-

al controls.25

Recommendations. When the diagnosis ofERD vs EE is not apparent despite endoscopy andiopsy, intraesophageal pH monitoring may be of use inxcluding pathologic reflux as either the primary or aoncomitant cause for esophageal eosinophilia (Grade B).lternatively, an upper endoscopy after 6 – 8 weeks ofigh-dose PPI treatment can help determine the etiologyf esophageal eosinophilia (see Treatment section).sophageal manometry does not provide diagnosticalue in patients with EE.

RadiographySome of the initial case series describing EE re-

orted esophageal narrowing.47,58,59 Since then, it is wellecognized that proximal and distal strictures are associ-ted with EE. In addition, long segment narrowing andecreased compliance of the esophagus have also beenescribed; these dynamic findings must be sought care-ully because they may not be apparent on routine study.chatzki ring has been described in some patients withE. In a series of 18 children with Schatzki ring, 8 were

ound to have EE.43 At endoscopy, none of these childrenhowed gross evidence of a ring, suggesting that theadiograph had shown a transient contraction. Thesendings suggest that narrowing observed at endoscopyay or may not be seen radiographically and vice versa.hen an esophageal contrast study is performed, close

ttention needs to be paid to esophageal distensibilitynd evidence of proximal or transient narrowing. In pa-ients with a history of chronic vomiting, upper GI series

ay be useful to investigate other possible anatomicauses of vomiting (eg, malrotation, hiatal hernia).

Recommendations. In patients with dysphagia,n upper GI contrast study may identify the presence of
stricture, as well as its caliber and length. A contrast b
tudy may be beneficial for children who present withomiting to rule out anatomic etiologies such as malro-ation (Grade C). This information is also potentiallyelpful for a subsequent upper endoscopy because it maylert the endoscopist to use a smaller caliber endoscoper to proceed particularly cautiously with passage of the

nstrument so as to lessen the likelihood of a mucosalear. In addition, it prepares the endoscopist for theossible need for a dilatation. An upper GI contrast study

s generally not useful in patients presenting with symp-oms typical of GERD, eg, heartburn.

HistopathologyHistory of Esophageal EosinophiliaIn 1977, the first report of eosinophilic inflamma-

ion of the esophageal epithelium in an adult with dys-hagia and no GERD symptoms was published.60 Overhe following few years, isolated case reports describeddditional similar findings in adults and children.61– 64

Throughout the 1980s, a number of reports associatedntraepithelial eosinophils in esophageal biopsy speci-

ens with GERD.65– 69 Interestingly, Leape et al67 andyams et al68 recognized that a number of patients whoad intraepithelial eosinophils in esophageal biopsy spec-

mens failed to respond to medical treatment for GERD.distinguishing feature of these patients was a dense

osinophilic infiltrate (�20 eos/HPF) in their esophagealucosa (see Figure 3). From 1982 until 1995, the signif-

cance of numerous esophageal eosinophils was underap-reciated. Most pathologists viewed the presence of in-raepithelial esophageal eosinophils as pathognomonicor GERD.

During this time, several investigators began to suggesthat GERD might not be the etiology in patients present-ng with a severe esophageal eosinophilia. In 1985, Leeeported a series of children and adults whose esophageal

igure 3. Eosinophilic esophageal inflammation in eosinophilic esophagi-is. (A) Low-power view of the epithelium showing increased numbers ofosinophils and evidence of basal zone hyperplasia and elongated reteapillae. (B) High-power view of the epithelium demonstrating large num-ers of eosinophils accumulating preferentially toward the luminal surfacelong with a thickened basal zone.

iopsy specimens showed “marked eosinophilia,” defined

aOpee1arapacts

cataalttdswrStHsweGa

iennnsmeiH

dtdm

nrmiprasalacat

nasn

aajabecipd

emp(

sooFi

T

N

AG

AIN

STITUTE


s �10 intraepithelial eosinophils in 2 HPFs (Table 5).70

ne patient, a 15-year-old girl presented with abdominalain, asthma, and peripheral eosinophilia and showed novidence of GERD. Lee considered this case to be anxample of “idiopathic eosinophilic esophagitis.” In993, 11 adults with dysphagia, normal pH monitoring,nd dense esophageal eosinophilia (�20 eos/HPF) wereeported.14 Importantly, control patients with GERD hadmean of 3.3 eos/HPF in their esophageal mucosa. Sevenatients had food hypersensitivity, and all requireddvanced intervention (dilation and/or steroids in onease) for resolution of symptoms. The authors cau-ioned about automatically attributing esophageal eo-inophilia to GERD.14

In 1995, a seminal article by Kelly et al reported 10hildren with GERD-like symptoms with intense esoph-geal eosinophilia despite antireflux therapy.23 Two ofhese patients had already received fundoplication, andll responded well to amino acid formulas, suggesting anllergic etiology. Subsequently, the degree of intraepithe-ial eosinophilic infiltration was correlated with responseo conventional GERD treatment in children.16,54 Duringhe 1990s, a number of studies described children withense esophageal eosinophilia (�15–20 eos/HPF) whohowed clinicopathologic response to dietary restrictionith an amino acid-based formula,23,54 oral corticoste-

oids,54,56,71 and topical corticosteroids.72 Additionally,teiner et al showed an inverse correlation between epi-helial eosinophil counts and reflux index, ie, 1–5 eos/PF correlated with an elevated reflux index.18 These

tudies provided additional confirmation that patientsith intractable GERD symptoms and dense eosinophilic

sophageal infiltrates appeared to have a unique non-ERD disorder, which in some cases seemed related to

llergy.

Quantitation of EosinophilsThe key diagnostic criterion for diagnosing EE

n all studies has been an increased number of intra-pithelial eosinophils. All studies used a thresholdumber of eos/HPF for the diagnosis of EE, but theumber and method used to generate that number wasot uniform. Peak count, the highest number of eo-inophils within a HPF, was the method most com-

only used.3,11,14,22,23,25,28,29,52,55,73 A mean number ofos/HPF was generated in some studies based on count-ng the number of eosinophils in several representative

able 5. Histologic Features Associated With EE

�15 Intraepithelial eos/HPF (peak count)Eosinophil microabscessSuperficial layering of eosinophilsBasal zone hyperplasia

OTE. None of the features are pathognomonic of EE.

PFs6,8,10,16,17,36,43 or in all HPFs.32,48,51,54,56 Most studies fi

id not report the magnification and/or dimensions ofhe HPF in which eosinophils were counted. Those thatid reported a wide variance in surface area, from 0.196m2 to 0.44 mm2.6,16,17,22,36

Number of Eosinophils That Define EEThe number of eos/HPF used to establish a diag-

osis of EE varied among studies. For example, 10 studiesequired �15 eos/HPF based on peak count7,23–25,45,73 or

ean number from a defined number of fields exam-ned6,17,43,56; 8 studies required �20 eos/HPF based oneak count3,14,15,22,28,52,55 or mean number10; 2 studiesequired �24 eos/HPF peak count29 or mean number32;nd 1 study required a peak count �30 eos/HPF.11 Onetudy took a novel approach and set threshold numberst �20 eosinophils in 1 HPF or �15 in 2 HPFs.8 Theowest number density of eosinophils reported for a di-gnosis of EE was 15 eos/HPF for either peak or meanounts. The maximal number of eos/HPF that is associ-ted with GERD-related esophagitis is still under inves-igation.2,14

Mucosal Biopsy Specimens From Other Partsof the GI TractIn most studies, investigators identified histologically

ormal biopsy specimens from stomach (generally antrum)nd duodenum.3,4,6–8,10,11,14–17,19,20,22–24,28,30,43,51,53–56 Fourtudies specifically excluded patients with eosinophilia ofonesophageal sites in their analysis of EE.10,11,56,73

Eosinophil Morphology and AssociatedHistopathologic Features Observed in EEDegranulation. Major basic protein has been used

s a marker for eosinophil degranulation in studies ofsthma and atopic dermatitis. Increased extracellular ma-or basic protein deposition in the esophageal mucosa ofdults with EE compared with those with GERD haseen reported.8,10 Other studies identified extracellularosinophil granules in the mucosa affected by EE, butontrols were not examined.8,14,25,28 A caveat in interpret-ng eosinophil degranulation is the fact that biopsyrocurement and processing may cause eosinophilegranulation.74,75

Microabscesses. Three studies determined thatosinophilic microabscesses, defined as aggregates of 4 orore eosinophils in a cluster, were found exclusively in

atients affected by EE and not in those with GERD8,10,16

see Figure 4).Superficial layering. Another histologic feature as-

ociated with EE is the preferential superficial distributionf eosinophilic inflammation in the upper one third to halff the squamous epithelium7,8,10,14,16,17,19,20,25,28,43,55 (seeigure 4). Surface layering was not found in biopsy spec-

mens from reference groups.10,16,55

Basal zone hyperplasia. Most investigators de-
ned basal zone hyperplasia as a basal zone that occupied

mFlclialrBwozppw

l

riip

eccsbnbaswat

FaascloaTesas

AG

AIN

STIT

UTE


ore than 20% of the epithelium8,10,14,16 –18,20,23,36,43 (seeigures 3 and 4). Some studies reported less prevalent or

ess marked basal layer hyperplasia in a reference groupompared with patients with EE.8,14,16,28,29,36,52 Papillaryengthening was variably defined and was reportedn 50%–100% of cases of EE in which it was evalu-ted.8,14,16,19,20,23,36 Some reports described papillaryengthening that was less prevalent8 or milder16,23 in aeference group compared with EE biopsy specimens.asal zone hyperplasia and papillary elongation requireell-oriented sections and therefore can be evaluated innly a minority of EE patients. Quantification of basalone hyperplasia can be accomplished by staining for theroliferating cell antigen Ki-67 (MIB1 antibody); indeed,atients with EE have increased Ki-67 staining comparedith reference control groups.21

Epithelial edema was occasionally described in histo-

igure 4. Eosinophilic microabscessssociated with eosinophilic esoph-gitis. (A) Low-power view of an eo-inophilic microabscess with superfi-ial layering of eosinophils along the

uminal surface. (B) Low-power viewf 2 massive eosinophil abscesseslong luminal border of esophagus.hese occur on a base of hyperplasticpithelium. (C) High-power view of eo-inophil microabscess. Inferior to thebscess are a number of eosinophils,ome of which appear degranulated.

ogic evaluations of EE,7,8 whereas epithelial ulcers were s

arely reported.8,16 Lamina propria fibrosis was describedn only 2 reports.8,36 A limitation in evaluating this find-ng is the absence of lamina propria in most esophagealinch biopsy specimens.

Other inflammatory cells. Cell types other thanosinophils were evaluated in some studies. Lympho-ytes were increased in EE biopsy specimens10,16,17,21,51

ompared with a reference group17,21,29,51 and in biopsypecimens obtained prior to therapy compared withiopsy specimens following therapy.17,21 Polymorpho-uclear leukocytes were reported in some studies of EEiopsy specimens8,14,16,48 and were more8 or less16,48

bundant in reference groups. Mast cells were de-cribed as scattered as opposed to aggregated14 andere increased in EE biopsy specimens compared withreference group.24,29,51 Numbers of mast cells posi-

ively correlated with numbers of eosinophils in one
tudy.29

espoPcondlecit(ntcspfi

misoebsrdGstst

thoEappIditiso

IetemoCaeetesreg

aca

gaeepewlrsmeocuttiospei

cters

s

AG

AIN

STITUTE


Recommendations. EE is a clinicopathologic dis-ase defined by esophageal symptoms associated with aevere isolated esophageal eosinophilia and absence ofathologic GERD as evidenced by normal pH monitoringf the distal esophagus or lack of response to high-dosePI treatment. Intraepithelial eosinophils should beounted in the most intensely inflamed HPF of the bi-psy (�400) to generate a peak count. Setting a fixedumber of eosinophils as the sole cut-off criterion toistinguish EE from GERD is contentious, possibly mis-

eading, and probably unwise based on current knowl-dge. On the basis of this literature review and collectivelinical experience, we conclude that a peak count of �15ntraepithelial eos/HPF is an absolute minimum numbero make the diagnosis of EE in the proper clinical contextGrade B). If all HPFs are counted, the mean eosinophilumber may be less than 15 because of focal inflamma-ion in the biopsy specimens, but at least 1 HPF mustontain at least 15 intraepithelial eosinophils. For re-earch purposes, defining EE with a higher threshold ofeak eosinophils may be advisable to increase the speci-city of the diagnosis.Additional features that are not pathognomonic butay be helpful to the pathologist in recognizing EE

nclude eosinophil microabscesses (correlate of endo-copic mucosal with specks and plaques), surface layeringf eosinophils, basal layer hyperplasia, papillary length-ning, degranulating eosinophils, and lamina propria fi-rosis and inflammation. These features should be as-essed in all biopsy specimens and included in pathologyeports in addition to the number of eosinophils. Theiagnostic criteria for adults are the same as for children.astroenterologists treating adults and children with

ymptoms of esophageal dysfunction and numerous in-raepithelial eosinophils in esophageal biopsy specimenshould ensure that the disease cannot be attributed solelyo GERD before making a diagnosis of EE.

Allergic EvaluationHistory, Physical Examination, and Testingfor Other Atopic DiathesesMost studies characterizing the allergic pheno-

ype have been performed in children. Allergic responsesave been strongly implicated in the etiology of EE basedn several lines of evidence. The majority of patients withE (50%– 80%)22 is atopic based on the coexistence oftopic dermatitis, allergic rhinitis, and/or asthma and theresence of allergic antigen sensitization based on skinrick testing or measurement of plasma antigen-specificgE. Importantly, most patients improve on allergen-freeiets, providing supportive evidence that antigen is elic-

ting the disease. Substantial evidence is accumulatinghat EE is associated with T helper cell (Th) 2 typemmune responses (the type of T helper cell polarizationeen in allergic individuals). In particular, elevated levels
f eosinophil-active Th2 cytokines (eg, interleukin (IL)-4, w
L-5, and IL-13) as well as mast cells are present in thesophagus of EE patients.5,24,51 In addition, experimen-al models of EE can be induced in mice by allergenxposure, especially in the respiratory tract followingucosal or epicutaneous sensitization, as well as by

verexpression of Th2 cytokines (IL-5 and IL-13).76 –79

ollectively, these experimental systems demonstraten intimate connection between the development ofosinophilic inflammation in the respiratory tract andsophagus not only in response to external allergicriggers but also to intrinsic Th2 cytokines. It is inter-sting to note that patients with EE sometimes reporteasonal variations in their symptoms; case reportsecently documented seasonal changes in esophagealosinophil levels, especially in the proximal esopha-us.80,81

Recommendations. Because of the high rate ofllergic rhinitis, asthma, and/or eczema in EE patients, aomplete evaluation by a well-informed allergist for othertopic diatheses is recommended (Grade C).

Assessment of Atopy by Analysis of BloodSamplesPeripheral eosinophil count and eosinophil

ranule proteins. Seven pediatric16,17,19 –21,26,47 and 4dult4,8,51,80 studies document the number of peripheralosinophils, the percentage of EE patients with periph-ral eosinophilia, and the levels of eosinophil granuleroteins. All were retrospective studies or case reportsxcept for one prospective cross-sectional study. Thereas a significant amount of variability in the defining

evel for “peripheral eosinophilia” (range of eosinophilseported as abnormal ranged from greater than 350 eo-inophils per mm3 to greater than 800 eosinophils per

m3). Some reports did not define the number of bloodosinophils that constituted a diagnosis of blood eosin-philia. Overall, 10%–50% of adults and 20%–100% ofhildren had elevated peripheral eosinophil counts butsually only modestly elevated (�2-fold). In all studies,here was a high percentage of concurrent allergic sensi-ization, and it is likely that concurrent allergic diathesesn conjunction with EE play a role in the elevated eosin-phil counts found in these patients. One study demon-trated that persistent blood eosinophilia correlated withersistent dysphagia.4 In another study, the degree oflevation of serum eosinophils correlated with the sever-ty of EE.82

Two studies document decreases in blood eosinophilounts following therapy. Following treatment with flu-icasone, 88% of patients demonstrated decreased bloodosinophil counts.26 In another study of oral corticoste-oids, most patients demonstrated decreased blood eo-inophils following treatment.56

Compared with other eosinophil products, plasma eo-inophil-derived neurotoxin (EDN) (but not stool EDN)
as also elevated in EE patients but had less predictive

vvwa7

btbtobamoletlssos

soArirtm7pdvcdl

mmlehs(ctfdrtdc

oagcEdeipp

atEapa

astotcitp7ttwat

aEet

adowolbassgPea

AG

AIN

STIT

UTE


alue than circulating eosinophil counts.82 When ele-ated levels of EDN and peripheral eosinophil countsere used together, sensitivity, specificity, and positivend negative predictive values were 63%, 92%, 83%, and9%, respectively.

Recommendations. Evaluation of peripherallood eosinophils may provide supportive evidence forhe presence of EE and the degree of tissue involvementut are not diagnostic, and correlation with disease ac-ivity is unknown (Grade C). In future studies, if eosin-phil levels are to be followed, it is important that (1)lood eosinophil levels be drawn at diagnosis and againt each evaluation for response to treatment (dietary oredical) and (2) notation is made regarding the control

f concurrent atopic diatheses and the extent of aeroal-ergen exposure at each time when eosinophil count isvaluated. Absolute eosinophil counts and defining cri-eria for “blood eosinophilia” should be reported in pub-ications that document peripheral eosinophilia. Furthertudies are needed to evaluate whether eosinophils con-titute an adequate surrogate disease marker either aloner in combination with other surrogate disease markersuch as EDN.

Total IgE. Five pediatric19,26,47,56,73 and 3 adulttudies4,51,83 report levels of total IgE in EE patients. Allf the studies are retrospective and/or case reports/series.s with peripheral eosinophil counts, the defining crite-

ia for abnormal values varied among studies, thus mak-ng broad conclusions difficult. One pediatric study cor-elated response of total IgE levels to oral corticosteroidherapy and found an almost 5-fold decrease, but the

ean IgE level remained above normal.56 Overall, 71%–8% of pediatric EE patients and 60%– 69% of adult EEatients had elevated total IgE levels. One adult studyemonstrates that peripheral IgE levels remained ele-ated for years in EE patients not undergoing pharma-ologic intervention.4 The high rate of concurrent atopiciatheses in these patients suggests that elevated IgE

evels were likely not linked specifically to EE.Recommendations. No published studies docu-

ent whether or not total IgE can serve as a surrogatearker for disease progression or resolution. If total IgE

evels are to be followed, it is imperative that (1) anvaluation is done regarding whether or not the patientas adequate aeroallergen avoidance and the pollen sea-on at each time when the total IgE level is evaluated and2) an evaluation is done regarding whether or not con-urrent atopic diatheses are adequately controlled at theime that the total IgE is evaluated. If IgE levels areollowed, it is recommended that levels be checked atiagnosis and at each endoscopic evaluation of diseaseesponse to intervention (Grade C). It is important thatotal IgE levels be interpreted within the context of age-efined normal values and that the total IgE level that is
onsidered “normal” be clearly stated in any publication. e
Aeroallergen-specific IgE. Although the presencef allergic rhinitis is cited in multiple studies, only onedult study specifically delineates the presence of anti-en-specific IgE to specific allergens (grass, a potentialross-reacting allergen to wheat and rye) in a patient withE.83 In addition, one case report suggests that EE isriven by aeroallergens,81 and one case series reportssophageal eosinophilia (up to 20 eos/HPF in the prox-mal esophagus and up to 12 in the distal esophagus) inatients with pollen allergies.80 Animal models have pro-osed that EE can be induced by aeroallergens.79

Recommendations. Given the high rate of otherllergic diatheses (50%– 80%) in EE patients and the po-ential of aeroallergens to have a role in the instigation ofE, it may be important to evaluate EE patients foreroallergen sensitivity (Grade C). Allergy testing mayredict the response to pharmacotherapy or dietaryvoidance in EE patients and thus warrants evaluation.21

Food-specific IgE. Three pediatric studies17,20,84

nd one adult study11 clearly delineate the results of foodpecific radioallergosorbent testing (RAST) in EE pa-ients. One study utilized partial elimination diet (with-ut success) based on food-specific IgE testing in addi-ion to skin prick testing.17 No other studies used alearly defined food specific-based elimination diet. Stud-es using empiric elemental formula or empiric elimina-ion diet in children without any allergy testing (skinrick and/or patch, antigen-specific IgE) documented a7%–98% disease improvement or eradication, suggestinghat any allergy testing utilized for foods may not iden-ify the inciting food allergen.55,85 It is currently unclearhy more patients with food-specific IgE do not havenaphylaxis to those foods for which they have positiveests.

Recommendations. There are no positive or neg-tive predictive values for food-specific IgE level testing inE. In vitro food allergy testing is not supported in thevaluation of EE patients at this time, and empiric foodesting should utilize skin prick tests (see below; Grade B).

Peripheral cytokines. Three studies, 1 adult51

nd 2 pediatric,29,82 evaluated peripheral cytokine pro-uction in EE patients. IL-13 release was elevated in 50%f adult patients (n � 3 patients with increased IL-13)ith EE as compared with controls; no differences in IL-5r interferon � were observed. In one study, eotaxin-3

evels were found to be elevated 2-fold in the peripherallood of EE (n � 12) as compared with normal (n � 6)nd chronic esophagitis patients (n � 5).29 In anothertudy involving 47 EE pediatric patients, eotaxin-3 washown to be elevated in EE and correlated with esopha-eal eosinophil levels (37.7 vs 11.5 pg/mL, respectively,� .01). In the same study, levels of plasma eotaxin-1,

otaxin-2, and IL-5 had no predictive value.82 Geneticnalysis of single nucleotide polymorphism (SNPs) in the
otaxin-3 gene demonstrated that SNP 2496 GG in the

3t

gitls

agwhtfmnus

apewae�septbe

eptrtea

thataatpsCtopa

afmrmmlTpea

att

d“icbiatsbhitol

mfaaidtctarpps

aTnttb

AG

AIN

STITUTE


=-untranslated region was overrepresented in EE pa-ients independent of atopic status.

Recommendations. Eotaxin-3 expression and itsenetic variation are promising markers of distinguish-ng EE from other causes of esophagitis (Grade B). Fu-ure research concerning the reversibility of eotaxin-3evels with therapy and their prognostic significance de-erve further investigation.

Gene expression. Two pediatric studies and onedult study evaluated changes in the level of esophagealenes in EE vs non-EE patients.24,29,51 Using a genome-ide transcript expression profile analysis, EE patientsave been demonstrated to have dysregulation of �1% ofhe human genome, with �50 genes changing over 10-old. In contrast, patients with chronic esophagitis were

uch more similar to normal individuals. In fact, aumber of these genes appear to be epithelial gene prod-cts, suggesting that the primary defect in EE may beecondary to an altered epithelial cell phenotype.

Using a genome-wide approach, followed by polymer-se chain reaction-based verification and assessment ofrotein elevations, eotaxin-3 has been identified as mark-dly elevated in EE patients (50- to 100-fold) comparedith normal individuals and those with chronic esoph-gitis. Although one study did not find elevated levels ofotaxin-3, the same study failed to find any gene elevated2-fold, suggesting technical limitations. Two studies

howed low or no difference in esophageal RANTESxpression; both showed increased esophageal IL-5 ex-ression.24,51 One study has shown increased levels ofumor necrosis factor � in affected tissue.51 There haveeen no demonstrable increases in cysteinyl leukotrienexpression in EE patients.30

Recommendations. Although the results ofotaxin-3 expression in EE vs non-EE patients are highlyromising, assessment of eotaxin-3 remains a researchool, and correlations with disease severity and activityemain to be evaluated (Grade B). The identified EEranscriptome may indeed have promising value for dis-ase diagnosis, assessment of therapeutic responsiveness,nd prognosis.

Skin prick testing for antigen sensitization. Fif-een studies involving 12 case series and 3 case reportsave examined skin prick testing in EE patients. Indults, positive skin tests to food allergens were difficulto elicit, except when there was a history of a reaction to

food (1 case). Positive skin tests to environmentalllergens were more frequently found than positive reac-ions to food antigens. In pediatric patients, more com-rehensive studies have been reported, including a retro-pective case series with a total of 786 patients.ollectively, these studies have shown that approximately

wo thirds of patients have positive skin tests to at leastne food allergen, whereas one third do not have anyositive skin tests. The number of foods tested were not
lways reported but varied from an average of 13 foods to m
panel of 42 foods. When larger panels were used, theoods tested included the common food allergens— cow

ilk, eggs, peanuts, soy, wheat, and fish—as well as rep-esentative members of classes of foods including grains,

eats, seafood, tree nuts, fruits, and vegetables. Theean number of positive skin tests to foods when the

arger panels were used varied from 2.7 � 3.3 to 6 � 4.2.he most common foods reported to be positive by skinrick tests included common food allergens—peanuts,ggs, soy, cow milk, and wheat—in addition to beans, rye,nd beef.

Recommendations. Skin prick testing for foodsnd environmental allergens should be considered sohat potential allergens and the atopic status of EE pa-ients are identified (Grade C).

Atopy patch testing in EE. Patch tests were firstescribed for contact dermatitis in the late 1890s forallergy” to fabric. The earliest publication on patch test-ng in eczema was described in 1937,86 with the earliestontrolled trial in 1982.87 Atopy patch testing (APT) haseen used for the diagnosis of non-IgE, cell-mediated

mmune responses in which T cells are thought to playprominent role. APT involves prolonged contact of

he allergen to the skin with the goal of mimicking aimilar immune response to atopic dermatitis. In fact,iopsy specimens of the patch test sites were found toave initial Th2 cell infiltration followed by a predom-

nance of Th1 cytokines and eosinophils88 similar tohe biopsy findings that have been observed in the skinf atopic dermatitis patients during acute and chronic

esions.89

APT has been most extensively studied in atopic der-atitis. Most studies find that APT was better in identi-

ying late reactions and GI reactions in children withtopic dermatitis.90 –92 APT has been studied primarily intopic dermatitis. The food to be tested is typically placedn aluminum cups (Finn Chambers on Scanpore; Aller-erm Laboratories, Inc. Petaluma, CA) and then appliedo uninvolved areas of the patient’s back in the 12-mmhambers.93 Similar to patch testing for contact derma-itis, a 48-hour occlusion time is used, and the patchesre subsequently read at 20 minutes and 24 hours afteremoval of the Finn chamber, examining for erythema,apules, and induration.94 Any food can be assessed withatch testing, although cow’s milk, hen’s egg, wheat, andoy have been studied most extensively.73

Application of APT in EEAPT has been used for the diagnosis of food

llergies in two published studies by Spergel et al.73,93

hey examined 146 children with biopsy specimen-diag-osed EE and eliminated foods based on positive skinest and atopy patch test. The authors found that 77% ofhe patients had resolution of their biopsy specimensased on these results (including 14% that required ele-
ental formulas because of the multiple positive food

aswScemtc

siIwswdrtidt(

lphottt

pPwabut

wecwawfltsiot

twstrpmt

aaitspt(ge

pTtmn

eeoaodw1ttpaVebaCteFtcpitAn

AG

AIN

STIT

UTE


llergies). Greater than 98% of their population re-ponded to an elemental diet,22 indicating that patientsho failed testing did not identify the correct foods.pergel et al also identified foods that were apparentlyausative based on reintroduction of single foods orlimination resulting in normalization of biopsy speci-ens on elimination or significant eosinophilia on rein-

roduction. The most common foods were milk, egg, soy,hicken, and wheat.73

Recommendations. The combination of prickkin tests and APT has been successful in one center ands being examined at other centers to verify these results.n addition, APT has shown promise in atopic dermatitisith good predictive values, high specificity, and low

ensitivity, and APT has shown highly promising resultsith regard to food elimination diet and food reintro-uction in patients with EE. However, its use should beeserved until additional data from multiple researcheams emerge that clearly establish its value for diagnos-ng and/or managing EE (Grade B). In addition, furtherata regarding the types of cells and immune responsehat is occurring at the site of patch testing are neededeg, skin biopsy studies).

Treatment of EEIt is not known whether treatment will impact

ong-term outcomes of the disease, and the exact endoints (reversal of symptoms and/or endoscopically oristologically normal mucosa) are not certain. The lackf evidence makes decisions regarding choice and dura-ion of treatment difficult. Here, we present the evalua-ion of the data regarding efficacy and safety of knownreatments.

Acid SuppressionRationale. Gastric acid is not thought to be the

rimary mediator associated with the pathogenesis of EE.atients with esophageal eosinophilia who are treatedith PPI with resolution of their symptoms have GERDnd not EE. Basic studies suggest that gastric acid inhi-ition may predispose to an allergic phenotype but pop-lation-based research has not yet been done to confirmhis hypothesis.95,96

Studies. PPIs play 2 potential roles for patientsith EE. First, they are useful as part of the diagnostic

valuation in patients suspected of having EE. Lack oflinicopathologic response to PPI treatment in patientsith an isolated esophageal eosinophilia is virtually di-gnostic of EE.18,19,22 On the other hand, some patientsith a well-established diagnosis of EE may develop re-ux symptoms intermittently that are responsive to PPIreatment. In a collection of 49 patients from 3 differenttudies, only 16% demonstrated significant symptomaticmprovement following PPI treatment.13,35,50 In a studyf 102 patients with esophageal eosinophilia who were
reated with acid blockade, those who presented with less i
han 3 eos/HPF improved clinically and histologically,hereas those with greater than 20 eos/HPF remained

ymptomatic and histologically abnormal.54,56 Rarely, pa-ients who have a significant esophageal eosinophiliaespond to PPI medication. Ngo et al reported that 3atients with significant esophageal eosinophilia (maxi-um of 37, 21, and 52 eos/HPF) responded dramatically

o PPI therapy.2

Recommendations. Acid suppression is useful aspart of fulfilling the diagnostic criteria for EE. In

ddition, it may be used in lieu of esophageal pH mon-toring for patients with established EE who have symp-oms secondary to concomitant GERD. PPI therapyhould not be considered as a primary treatment foratients with EE. Rather, it may be considered as co-herapy because it sometimes alleviates symptoms in partGrade C). It is interesting to speculate that the esopha-us of EE patients may have enhanced sensitivity to acid,ven in the absence of pathologic reflux.

Esophageal DilatationRationale. A number of studies document the

resence of esophageal narrowing in patients with EE.he incidence of this complication is not certain. By the

ime this complication arises, medical management aloneay not suffice, and thus mechanical dilatation may be

ecessary.Studies. Several studies have reported the use of

sophageal dilatation of EE strictures in adults. Morrowt al described 19 patients who underwent dilatation: 15f 16 reported overall improvement in their dysphagiafter multiple sessions of dilatation.50 No perforationsccurred; however, deep mucosal tears, increased posten-oscopy analgesia, and difficulty inserting the endoscopeere reported in several patients. Straumann et al studied1 EE patients who required esophageal dilatation.4 Ofhese, only 4 patients required repeat dilatation. Al-hough the procedure caused severe mucosal tearing, noerforations occurred. Over 50% became asymptomatic,nd one patient reported no improvement in symptoms.asilopoulos et al described 5 patients with small calibersophagus.39 Of these patients, all received esophagealougienage; 2 experienced extensive esophageal tearingssociated with chest pain and overnight hospitalization.antu et al reported successful esophageal balloon dila-

ation in 2 cases.97 In the only pediatric study, Nurkot al described 7 EE patients who underwent dilatation.ive of the 7 patients had total symptom relief, whereashe other 2 only had a partial response.43 Despite suc-essful esophageal dilatation, a significant number ofatients developed a recurrence of their stricture requir-

ng repeat dilatation. The recurrence rate ranged from 7%o 50% and occurred between 2 and 24 months.4,36,50,98

dditionally, as mentioned above, there have been aumber of reports of esophageal mucosal tearing, signif-

cant pain, and rare reports of esophageal perforation.9

Met

nistsatppspTntWbcmtud

factpmptdet

sbPelflsgcnsvwsst

dacdnd2co

rThtppwtsmlcidefTc

tthtssbpta

sapslidrartfp

AG

AIN

STITUTE


oreover, there has been an association with linearsophageal renting or tearing simply with the introduc-ion of the endoscope through the stricture.

Although dilatation may not alter the underlying ab-ormal esophageal histology, it may be required to facil-

tate esophageal function. Whether medical therapyhould always be considered prior to performing dilata-ion of strictures secondary to EE is not certain. Notudies have demonstrated normal esophageal histologyfter dilatation without additional medical or dietaryherapy. Unless a critical stricture exists, a diagnosticrocedure should be performed. If esophageal eosino-hilia exists, patients not previously treated with aciduppression should be started on PPI therapy. If EEersists, medical or dietary therapy should be initiated.he approach to the untreated patient with narrowing isot certain. Approaches include pretreatment with nutri-ional or medical treatments or immediate dilatation.

hether medical treatment before dilatation leads to aetter outcome is not known. Although not proven, con-ern exists that the use of steroids (systemic or topical)ay exacerbate the risk of a perforation related to dila-

ation. Expert opinion suggests that residual strictures,nresponsive to medical therapy, may be more safelyilated, thereby reducing the risk of esophageal tearing.

Recommendations. Esophageal dilatation is use-ul for symptomatic patients who present with symptom-tic esophageal narrowing secondary to fixed stricturesausing food impaction (Grade C). However, because ofhe risk of mucosal tearing and perforation, wheneverossible, a diagnostic endoscopy with biopsy followed byedical or dietary therapy for EE should be attempted

rior to performing esophageal dilatation. Inspection ofhe esophageal mucosa (radiographic or very gentle en-oscopic examination) should be considered followingsophageal dilation to assess for laceration injury prior tohe performance of sequential, larger caliber dilation.

CorticosteroidsRationale. Eosinophilic inflammation acutely re-

olves with the use of systemic corticosteroids in a num-er of allergic diseases including asthma and eczema.roposed mechanisms in which corticosteroids impactosinophils include induction of apoptosis, down-regu-ation of chemotactic factors, and inhibition of proin-ammatory mediators. As such, a number of studies havehown that corticosteroids significantly improve esopha-eal eosinophilia in patients with EE. Although systemicorticosteroids are effective, they are associated with sig-ificant adverse effects. In contrast, swallowed topicalteroids administered by a metered dose inhaler (or in aiscous solution) provide several advantages comparedith systemic steroids. The dose of topical steroid is

ignificantly less, the liver rapidly metabolizes the topicalteroid, and the delivery of the medication is directly to
he esophageal mucosa. i
Studies: systemic steroids. Two early studies haveemonstrated that systemic glucocorticoids (prednisone)re an effective pharmacologic treatment in resolving thelinicopathologic features of EE.61,71 In 1998, Liacouras et alemonstrated that the use of systemic corticosteroids sig-ificantly improved both clinical symptoms (within 7ays) and esophageal histology (within 4 weeks) in 20 of1 children with EE.56 No published studies to date haveompared the impact of systemic corticosteroids withther treatments.

Clinical experience dictates that systemic corticoste-oids are useful when urgent symptom relief is required.hese patients include those with severe dysphagia, de-ydration, significant weight loss, or esophageal stric-ures. Additionally, steroids may be useful for EE patientsresenting with a small caliber esophagus or for thoseatients deemed at high risk for esophageal perforationhen undergoing esophageal dilatation. Dosages effec-

ive in relieving clinicopathologic abnormalities wereimilar to those used for inflammatory bowel disease (1–2

g/kg/day of prednisone; maximum 60 mg), althoughower doses have not been reported.56 Risk factors asso-iated with long-term use of systemic corticosteroidsnclude growth abnormalities, bone abnormalities, moodisturbances, and adrenal axis suppression among oth-rs. Corticosteroids were weaned similar to a scheduleollowed for patients with inflammatory bowel disease.ypically, when the medication was discontinued, thelinicopathologic signs and symptoms recurred.

Studies: topical steroids. Beginning in 1998, mul-iple studies demonstrated the effectiveness of swallowedopical corticosteroids delivered from a metered dose in-aler unit in treating clinical symptoms and abnormal his-ology associated with EE in adults and children. The firstuch study was reported by Faubion et al who prescribedwallowed fluticasone propionate (up to 880 �g/day) oreclomethasone twice a day to 4 patients with EE.72 All 4atients demonstrated an improvement in clinical symp-oms; 1 patient underwent repeat posttreatment biopsynd showed resolution of mucosal eosinophilia.

Since that initial study, 47 adults and 33 children weretudied in 4 separate studies.6,15,17,21 With regard to thedults studied, 440 –500 �g twice daily of fluticasoneropionate was administered for 4 – 6 weeks. Clinicalymptoms improved in all but 1 patient; complete reso-ution occurred in 75% of cases. Reported adverse effectsncluded esophageal candidiasis in 3 patients and severery mouth in 1 patient. Follow-up evaluation revealed aecurrence in symptoms in 17 of 37 patients between 3nd 18 months after therapy was discontinued. Withegard to the pediatric patients studied, 220 – 440 �gwice daily of fluticasone propionate was administeredor 6 –12 weeks. Clinical and histologic symptoms im-roved in 31 of 33 patients; 2 patients had no significant

mprovement. Esophageal candidiasis developed in 6 pa-

ta

wdtasacooan

fdntpesd

stgiiasmrtasg

taflhneowosdhtmtMls

dfMcsm

tgct

mtsmdo

satdAstidtwoomEi

htcasMgdc

dpaf

AG

AIN

STIT

UTE


ients. Long-term follow-up and recurrence of symptomsfter therapy was discontinued was not reported.

Details regarding the exact method of administrationere not always presented, but clinical experience andocumented protocols recommend that patients sprayhe metered dose inhaler in the mouth with lips sealedround the device. Following administration, patientshould not eat, drink, or rinse for 30 minutes.17,81 In anttempt to provide easier delivery of this form of medi-ation, Aceves et al used a preparation of viscous budes-nide.99 Two patients were studied and swallowed 500 �gf oral budesonide mixed in a sucralose suspension twiceday. The patients’ symptoms as well as histopathologyormalized within 3 months of initiating therapy.A meta-analysis of the risk of inhaled corticosteroids

ound that fluticasone might lead to bone loss at totalaily doses higher than 750 �g each day. This risk mayot be as great with swallowed topical steroids becausehey are rapidly metabolized by the first pass effect notresent with inhaled steroids.100 Although this studyvaluated inhaled steroids in patients with asthma, ithould be noted that total daily doses of 1760 �g eachay have been reported in patients with EE.15

Recommendations. Systemic and topical cortico-teroids effectively resolve acute clinicopathologic fea-ures of EE; however, when discontinued, the diseaseenerally recurs. Systemic corticosteroids may be utilizedn emergent cases such as dysphagia requiring hospital-zation, dehydration because of swallowing difficulties,nd weight loss. However, because of the potential forignificant toxicity, their long-term use is not recom-

ended (Grade B). For many patients, topical corticoste-oids are also effective in inducing EE remission. Al-hough the incidence of adverse effects with this form ofdministration has not been formally studied, severaltudies have documented its safety, except for local fun-al infections.

The use of topical corticosteroids for maintenancereatment has not been studied. Age adjusted doses anddministration frequency of topical corticosteroids, ie,uticasone, budesonide, for children and adults with EEave not been established and these formulations wereot designed for esophageal administration. One studyxtrapolated doses from those utilized in the treatmentf asthma.17 Since then, others have utilized higher dosesithout significant complications.6,15,21,116 On the basisf expert opinion and the current literature, suggestedtarting doses range from 440 – 880 �g per day for chil-ren and 880 –1760 per day for adolescents/adults. Drugas been administered by mouth and can be split intowice or 4 times daily doses. Equally important is the

ethod of administration; patients should be instructedo administer the MDI without the use of a spacer. The

DI should be inserted into the mouth, sprayed withips sealed around the device. The powder should then be
wallowed and not rinsed. Patients should not eat or o
rink for at least 30 minutes. This regimen is continuedor 6 – 8 weeks and then patients followed as described in

onitoring section (Grade B). More studies are needed tolarify specifics of topical steroid treatment plans. Alsoee Update section for information on alternative

ethod of administration.

Leukotriene Receptor Antagonists and MastCell StabilizersRationale. Inflammatory mediators such as cys-

einyl leukotrienes or preformed mediators found inranules released by mast cells have been theorized toause esophageal inflammation and tissue eosinophiliahat occurs in patients with EE.

Studies: cromolyn sodium. The use of oral cro-olyn sodium has never been formally studied in pa-

ients with EE. In a 10-year review, Liacouras et al pre-ented information on 14 EE patients treated with 100

g oral cromolyn, 4 times daily for 1 month.22 The studyemonstrated that no patient improved either clinicallyr histologically.

Studies: leukotriene receptor antagonists. Twotudies have addressed the role of leukotriene receptorntagonists in patients with EE. Attwood et al utilized upo 100 mg in 8 patients diagnosed with EE manifested byysphagia and symptoms of gastroesophageal reflux.13

fter several weeks of treatment, 7 of the 8 patientshowed complete symptomatic resolution; the other pa-ient improved but did not completely resolve. The med-cation was maintained for a median of 14 months atoses ranging from 20 to 40 mg per day. Once discon-inued, 6 of 8 patients had a recurrence of symptomsithin 3 weeks. Minimal adverse effects (nausea, myalgia)ccurred; however, no significant improvement in histol-gy was appreciated. Gupta et al determined esophagealucosal levels of cysteinyl leukotrienes in children with

E and normal controls and found that they were similarn both groups.30

Recommendations. Although cromolyn sodiumas no significant adverse effects, it has no apparentherapeutic effect for patients with EE. Leukotriene re-eptor antagonists have been shown to induce symptom-tic relief at high dosages; however, its use has not beenhown to have any effect on esophageal eosinophilia.

easurements of mucosal leukotriene levels do not sug-est potential for a therapeutic benefit. The use of theserugs for the treatment of EE is not supported by theurrent literature (Grade C).

Dietary TreatmentRationale. There is strong circumstantial but not

efinitive evidence that food allergens contribute to theathogenesis of EE in children. The removal of foodntigens has clearly been demonstrated to treat success-ully both the symptoms and the underlying histopathol-
gy in the majority of patients with EE. The elimination

omatdaat

ttsvflstttttcEart

aet(tcccspnapmrng

adbeetshwnia

fif

aswdsrptf

tlmtdaipm

nhctma

umf

tgatipmr

cmhioac

AG

AIN

STITUTE


f causative foods can follow several therapeutic regi-ens. First, specific food elimination can be based on

llergy testing and clinical history. Second, the decisiono limit foods can be based on simply removing the foodseemed to be the most likely to cause EE. Finally, anmino acid-based formula can be utilized, thus removingll potential food allergens. The effectiveness of dietaryherapy in adults has not been studied.

Studies: specific food elimination. As was men-ioned in the section entitled Allergy Testing, diagnosticests consisting of radioallergosorbent testing and IgEkin prick tests are limited and may have poor predictivealue in implicating those antigens that provoke an in-ammatory response in the esophageal mucosa. Severaltudies have demonstrated poor correlation of these testso improvement of either symptoms or tissue inflamma-ion.17,21 However, in one academic center, the introduc-ion of APT used in combination with IgE skin prickesting significantly increased the ability to identify po-ential food allergens. Spergel et al described the use ofombination skin prick and APT in 146 patients withE.73 Of these, 112 patients (77%) demonstrated clinicalnd histologic improvement after 6 weeks of dietaryestriction based on allergy testing utilizing skin prickests and APT.

Studies: removal of selected causative foods. Inn attempt to minimize allergy testing and determinetiologic food allergens, Kagawalla et al demonstratedhat the removal of the 6 most common allergenic foodsdairy, eggs, wheat, soy, peanuts, fish/shellfish), withouthe aid of allergy testing, demonstrated significant effi-acy.85 Approximately 74% of the 35 patients who re-eived the 6-food elimination diet demonstrated signifi-ant improvement both clinically and histologically. Thetudy also compared the 6-food elimination diet to 25atients who received a strict amino acid-based diet witho other added foods. The comparison revealed that,lthough both diets significantly improved the clinico-athologic features of the disease, the elemental diet wasore effective with regard to the number of patients who

esponded (22 of 25) and with regard to the residualumber of eosinophils per HPF (13.6 in the 6-foodroup; 3.7 in the amino acid formula group).

Studies: amino acid-based formula. The use ofn amino acid-based formula is currently the gold stan-ard in determining whether food antigens are responsi-le for EE in those patients who do not respond to dietlimination of specific antigens. In children, the use of anlemental formula has been shown to be extremely effec-ive in 92%–98% of patients.23,55 Patients’ symptoms re-olved within 7 to 10 days followed by almost completeistologic resolution of the esophageal eosinophiliaithin 4 to 5 weeks. After the symptoms and histologyormalized, a slow reintroduction of select foods was

nitiated. Because of poor palatability, the use of a strict
mino acid-based formula frequently required enteral a
eeding via a nasogastric or gastrostomy tube as reportedn these studies. Additionally, the administration of theseormulas can be costly.

Recommendations. Dietary therapy (the specificntigens removal or elemental formula) should be con-idered as an effective therapy in all children diagnosedith EE (Grade B). When deciding on the use of a specificietary therapy, the patient’s lifestyle and family re-ources also need to be considered. Consultation with aegistered dietitian is strongly encouraged to ensure thatroper calories, vitamins, and micronutrients are main-ained. The use of dietary therapy in adults requiresurther study.

BiologicsRationale. Potential future treatment includes

he use of monoclonal antibodies such as anti-IL-5. Bio-ogic treatments, like anti-IL-5, specifically target the

olecule receptors that influence the production, migra-ion, and activation of the eosinophil, subsequently re-ucing esophageal tissue inflammation.101 Notably,nti-IL-5 or IL-5 gene deletion blocks induction of exper-mental EE in mice, including esophageal epithelial hy-erplasia, providing preclinical evidence that this strategyay be effective in EE.

Studies. Garrett et al demonstrated the effective-ess of anti-IL-5 antibodies in an adolescent male whoad hypereosinophilic syndrome manifested by signifi-ant esophageal eosinophilia.102 Anti-IL-5 was adminis-ered via an intravenous infusion monthly. Within 3

onths of treatment, the patient’s symptoms and esoph-geal eosinophilia dramatically improved (Grade C).

Recommendations. Novel biologic agents present anique opportunity for certain patients with EE. Theseolecules await clinical trials and cannot be recommended

or routine use at the present time (Grade C).

Treatment Panel DiscussionTreatment recommendations are based on the po-

ential unknown deleterious impact of chronic esopha-eal eosinophilia. Minimizing treatment adverse effectsnd maintenance of a high quality of life are additionalreatment goals. Drug treatments are less restrictive, plac-ng no compromises on the patient’s diet, but carryotential adverse effects and unknown duration of treat-ent. Dietary treatments give the prospect of prolonged

emission but entail significant lifestyle modification.Despite the advances that have been made, controversy

ontinues regarding treatment end points. Should treat-ent be aimed at symptomatic improvement or toward

istologic resolution of eosinophilia? An analogy withnflammatory bowel disease may be made: the treatmentf inflammatory bowel disease and celiac disease is aimedt symptomatic relief and not histological normalcy. Inontrast, chronic esophageal inflammation secondary to
cid reflux has been shown to contribute to Barrett’s

edaobr

maaHoaathgsatbptltmna

iEcmblu

iebpsi

tucamt

eeawi

igso(iaIe

csiSrd

Eapslat

maptebtslbppt

efie2rtttu

icgn

AG

AIN

STIT

UTE


sophagus and potential esophageal cancer. Because weo not yet know the impact of chronic, untreated esoph-geal eosinophilia, there is an urgent need to determine,r even estimate, the natural history of EE so that we canetter weigh the morbidity of chronic therapy against theisk of future complications.

Treatment end points. The goal of therapy re-ains unsettled. At the minimum, treatment should be

imed at relieving symptoms, which would ideally beccompanied by resolution of esophageal eosinophilia.owever, resolution of symptoms and esophageal eosin-

philia may not occur concordantly. Substantial esoph-geal eosinophilia persists in some patients who aresymptomatic or have only minimal symptoms. The op-imal approach to such patients is unclear. On the oneand, more aggressive treatment may help prevent pro-ression to permanent esophageal dysfunction and pos-ibly improve symptoms that may not have been fullyppreciated by the patient or parents. On the other hand,he natural history of esophageal eosinophilia has noteen well established, and it is thus not clear whether allatients with persistent esophageal eosinophilia are des-ined to develop complications. There are no well-estab-ished markers to predict patients at increased risk, al-hough those who have already developed esophageal

orphologic abnormalities (such as rings, strictures, orarrowing) have already established themselves as beingt increased risk.

At the same time, more aggressive treatment may havemplication on quality of life and/or adverse effects.limination of foods in children, for example, can de-rease quality of life in the patients and their familyembers. Corticosteroids, even if given topically, have

een associated with esophageal candidiasis, and theong-term safety of strategies involving such treatment isnknown.Thus, we suggest that treatment be initially aimed at

mproving symptoms. In those with persistent esophagealosinophilia, the decision to advance treatment should beased on the degree of symptoms, the age of the patient, theresence of esophageal morphologic abnormalities, the re-ults of monitoring (see below), and the patient’s and fam-ly’s values and preferences (Grade C).

Monitoring. Optimal strategies for monitoring pa-ients with EE have not been established, and there contin-es to be variability among experts. EE has not been asso-iated with the development of esophageal malignancy,lthough follow-up has been short. Thus, the main aim ofonitoring is to prevent progressive esophageal dysfunc-

ion and detect complications from therapy.In children with EE, many experts perform periodic

ndoscopy and/or barium studies to evaluate for persistentsophageal eosinophilia and/or the development of esoph-geal morphologic abnormalities. However, as noted above,hether the detection of persistent esophageal eosinophilia

n asymptomatic patients warrants more aggressive therapy a

s unclear. Furthermore, there have been few studies touide the biopsy protocol and the optimal interval forurveillance and to clarify baseline fluctuation in the densityf esophageal eosinophilia with time. Whether surveillanceand the subsequent response to findings on surveillance)mproves outcomes is unclear. Surveillance endoscopiesre also associated with the potential for complications.n particular, patients with EE are at increased risk forsophageal tears and perforation.

Similar issues apply to adults with EE. Adults are moreapable than younger children to report accurately theirymptoms; however, whether monitoring symptoms alones sufficient to guide treatment decisions remains uncertain.ome experts suggest periodic, regular, upper endoscopyegardless of symptoms, whereas others suggest upper en-oscopy guided mainly by changes in symptoms.

Recommendations. In children and adults withE, we suggest regular clinic visits during which the patientnd parents should be questioned about symptoms, com-liance with therapy, and adverse effects (Grade C). Thisuggestion is based on improving the recognition ofong-term complications associated with chronic esoph-geal eosinophilia; presently, the incidence of complica-ions is unknown.

In children, options for endoscopic and radiographiconitoring should be discussed considering the issues

bove. One approach might be to perform repeated up-er endoscopies until settling on a treatment regimenhat has controlled symptoms and ideally resolvedsophageal eosinophilia. Repeat examinations can beased on change in symptoms or compliance withherapy. If repeat endoscopy with biopsy is planned, ithould be performed no sooner than 4 weeks after theast therapeutic intervention. These suggestions areased on improving the recognition of long-term com-lications associated with chronic esophageal eosino-hilia; currently, data are not available to determinehe optimal method to follow patients.

In asymptomatic patients with persistent esophagealosinophilia, a repeat upper endoscopy can be performedollowing institution of additional treatment. For thosen whom additional treatment is deferred, a repeat upperndoscopy and/or barium swallow can be obtained every

to 3 years to evaluate for progressive disease, but theisks of this approach outside of a clinical research pro-ocol need to be weighed against the unknown benefits;his is especially important because the accuracy of his-ologic and radiologic predictors of disease progression isnclear.The approach to adults with EE should consider sim-

lar principles as described above for children. However,linical experience suggests that adults may be inclined touide treatment based mainly on symptoms. Thus, theeed for surveillance should also consider willingness to
ccept more aggressive treatment based on the results.

crfmdsdtrdgmg

isorpemfnatid

atspaitbtgmtiat

rtwcitni

2PateNiaAdttv

Erttrpcphiw

eremshasdiidtom

T

AG

AIN

STITUTE


Future ResearchToday, the care of patients with EE stands at a

rossroad (see Table 6). Clinical experience and the cur-ent literature dictate that EE is a chronic disease withew patients, if any, outgrowing their illness. Whether

erely a subset or a majority of these patients is at risk toevelop irreversible fibrotic changes is not known. If atricture develops, the timetable is also unknown. Theseilemmas complicate the question: What are appropriatereatment end points: symptom resolution, histologicemission, or both? Natural history studies will be key toetermining whether treatment alters the disease pro-ression. Basic studies directed at understanding theechanisms by which eosinophils contribute to esopha-

eal injury will be critical.Currently, clinicians are limited to endoscopic biopsies

n making the proper diagnosis and, in many circum-tances, establishing disease remission. The identificationf novel methods to assess disease activity, by reducingeliance on repeated endoscopies, will significantly im-rove the quality of life of patients and likely reduce theconomic burden of the disease. Mechanistic pathwaysust be further delineated to identify target molecules

or intervention. Results of these studies will also lead toovel therapeutic approaches (such as anti-IL-5 antibody,nti-eotaxin-3 antibody) that will be particularly impor-ant for patients who cannot comply with current med-cal and nutritional approaches, those with recalcitrantisease, and those with fibrotic changes.Finally, instituting appropriate treatment has becomechallenging and time-consuming clinical process. Nu-

ritional (dietary antigen elimination, elemental diet) andteroid treatments are both effective in inducing clinico-athologic remission in most patients, but some patientsre still recalcitrant or encumbered by nonadherence orntolerable adverse effects. The definition and recogni-ion of disease subtypes, using clinical phenotypes and/oriomarker profiles, holds promise for future identifica-ion of at-risk (for fibrosis and stricture formation) sub-roups for more aggressive therapeutic intervention andonitoring. For instance, recent studies suggest that

here are allergic and nonallergic phenotypes, and thisnformation may guide future treatment into differentpproaches. The challenge will be to develop appropriate

able 6. Unresolved Issues

Maintenance medical managementTreatment end pointsNatural historyBest method to identify food/aeroallergensDegree of eosinophilia associated with GERDEvaluation and management of the asymptomatic patient with

esophageal eosinophiliaEtiology and pathogenesis

ranslational studies that define valid phenotypic subsets. t

The past decade has witnessed the emergence andecognition of a new disease entity with unique featureshat differentiate it from gastroesophageal reflux disease,ith which it had been confused. The joint efforts of

linical scientists and basic scientists to undertake stud-es on natural history, pathophysiology, biomarkers, andherapeutic approaches will be critical to developingovel diagnostic tools and therapeutic options that will

mprove the lives of affected children and adults.

Update Since the First InternationalGastrointestinal Eosinophil ResearchSymposiumAt the time of acceptance of this article (July

007), there has already been a remarkable 25% growth ofubMed articles concerning EE including high-qualityrticles about epidemiology, diagnosis, pathogenesis, andreatment (including the first controlled clinical trial andarly results with a novel targeted biotherapeutic agent).otably, these studies were contributed by investigators

n 3 continents, including Europe (Sweden, Switzerland,nd Spain), America (United States and Canada), andustralia; this highlights the expanded global health bur-en of EE. Interestingly, a recent comprehensive litera-ure review about the diagnostic criteria for EE indicatedhe need for a consensus standard,103 highlighting thealue of our current report.

In the area of diagnostics, the striking association ofE with food impaction,104 dysphagia,105 and multi-inged esophagus was reported in several papers.105 In-erestingly, the familial association of EE was expandedo include the cooccurrence of dysphagia and Schatzkiings in family members of EE patients.106 Furthermore,reliminary studies concerning cooccurrence of EE witheliac disease, erythema nodusum, and anticonvulsant hy-ersensitivity syndrome were reported.107–109 One long-termistory study in pediatrics reported the chronic and relaps-

ng nature of EE and a strong predominance amonghites.110

In the area of diagnostics, an important surveillancendoscopy study in 1000 healthy individuals in Swedenevealed the presence of esophageal eosinophilia in 5% ofndoscopic biopsy specimens, with 1% of individualseeting diagnostic criteria for EE.111 This raises the pos-

ibility that the prevalence of EE may be substantiallyigher than expected, especially in view of the chronicnd relapsing nature recently reported.110 A retrospectivetudy reported that the incidence of EE in a pathologyatabase in Iowa was similar in 1990 and 2005, suggest-

ng that the recent recognition of EE is not due toncreased disease awareness rather than increased inci-ence.112 In addition, 2 early studies reported the poten-ial utility of noninvasive biomarkers based on the levelsf peripheral blood eosinophils, EDN, eotaxin-3, andononuclear cell Th2 cytokine production in response

o food and/or aeroallergens.82,113

rsaalvdtiedaptcppiti

yrtofpdriacobp8to

sawctapefSpEce

aHi

ot

NC

U

oS

tv

tH

nUO

tpD

It

Sdn

FMn

M

PG

N

AG

AIN

STIT

UTE


Several major advances in the area of therapy wereeported, including an early provocative study demon-trating the ability of a humanized anti-IL-5 monoclonalntibody markedly to reduce esophageal eosinophiliand improve clinical symptoms in adult patients withong-standing disease.114 Positive and negative predictivealues for skin testing and patch testing for EE wereetermined for the most common foods in EE.115 Fur-hermore, the first double-blind, placebo-controlled trialn EE was reported.116 In particular, the investigatorsvaluated the effect of swallowed fluticasone (880 �gaily) vs placebo in a 3-month trial. Notably, they wereble to demonstrate strong efficacy of fluticasone com-ared with placebo, but only 50% of patients respondedo the study drug, likely because of the existence oforticosteroid resistance in a substantial number of EEatients. In addition, the investigators defined the firstlacebo effect in EE at �10%; this will have significant

mpact on the design of future clinical trials. A retrospec-ive study reported the success of oral viscous budesoniden 20 pediatric patients.99

In a retrospective study of 20 children, median age 4.1ears (1.7–17.6), some 85% responded within 3– 4 monthsesolution or improvement of symptoms and histopa-hology to use of topical steroid administered as a slurryf oral viscous budesonide (OVB). They had previouslyailed PPI treatment alone. This form of treatment isarticularly suitable for younger children who may haveifficulty with taking medication by inhaler. Patientseceived OVB 1–2 mg daily and were instructed not tongest any solid or liquid food for 30 minutes after itsdministration. Children under the age of 10 years re-eived OVB 1 mg daily and those who were 10 years orlder received 2 mg/day. Viscous budesonide was madey mixing each 0.5 mg Pulmicort Repulse with 5 g (5ackets) or sucralose (Splenda) to create a volume of–12 mL. A Pulmicort Respule is liquid budesonide in-

ended for nebulized administration (0.5 mg budes-nide/2 mL).

Finally, advances in understanding disease pathogene-is included an important paper describing the role ofdaptive immunity in a murine model of EE (consistentith the view that EE is an antigen-driven disease) and a

ompelling paper that demonstrated the presence of ex-ensive tissue remodeling (including deposition of collagennd evidence of angiogenesis) in EE biopsy specimens fromediatric patients.117,118 Of note, the investigators providedvidence for the overexpression of transforming growthactor � and its signaling molecule, phosphorylatedMAD2. The level of esophageal remodeling as well as theresence of activated IgE-bearing mast cells distinguishedE from GERD patients.119,120 Additionally, evidence forooperation between systemic Th2 immunity and local
otaxin-3 production was reported.121 C
Appendix 1. FIGERS Subcommittees

Clinical Features and Diagnostic TestingChair: Sandeep K. Gupta, MD, Division of Pedi-

tric Gastroenterology, Hepatology, and Nutrition, Rileyospital for Children, Indiana University School of Med-

cine, Indianapolis, IndianaSamuel Nurko, MD, MPH, Division of Gastroenterol-

gy, Hepatology, and Nutrition, Children’s Hospital Bos-on, Harvard Medical School, Boston, Massachusetts

Nirmala Gonsalves, MD, Division of Gastroenterology,orthwestern University, Feinberg School of Medicine,hicago, IllinoisPeter Bonis, MD, Division of Gastroenterology, Tufts

niversity School of Medicine, Boston, MassachusettsAlex Straumann, MD, Department of Gastroenterol-

gy, University Hospital Basel, University Basel, Basel,witzerland

Jonathan Markowitz, MD, MSCE, Pediatric Gastroen-erology and Nutrition, Greenville Hospital System, Uni-ersity Medical Center, Greenville, South Carolina

HistopathologyChair: Margaret H. Collins, MD, Division of Pa-

hology, University of Cincinnati, Cincinnati Children’sospital Medical Center, Cincinnati, OhioSimon P. Hogan, PhD, Division of Allergy and Immu-

ology, Cincinnati Children’s Hospital Medical Center,niversity of Cincinnati College of Medicine, Cincinnati,hioGlenn T. Furuta, MD, Section of Pediatric Gastroen-

erology, Hepatology, and Nutrition, The Children’s Hos-ital, Denver, University of Colorado Medical School,enver, ColoradoDon Antonioli, MD, Department of Pathology, Beth

srael Deaconess Hospital, Harvard Medical School, Bos-on, Massachusetts

Eduardo Ruchelli, MD, University of Pennsylvaniachool of Medicine, Department of Pathology, The Chil-ren’s Hospital of Philadelphia, Philadelphia, Pennsylva-iaHector Melin-Aldana, MD, Northwestern University,

einberg School of Medicine, Pathology and Laboratoryedicine, Children’s Memorial Hospital, Chicago, Illi-

oisMargret S. Magid, MD, Department of Pathology,ount Sinai School of Medicine, New York, New York

EndoscopyChair: Chris Justinich, MD, FRCPC, Division of

ediatric Gastroenterology, Queen’s University, Kingstoneneral Hospital, Kingston, Ontario, CanadaIkuo Hirano, MD, Division of Gastroenterology,orthwestern University, Feinberg School of Medicine,
hicago, Illinois

tv

oP

eP

ol

AMi

nsP

ovO

sC

Nd

Cv

ePt

rnC

pB

daW

M

en

tc

AG

AIN

STITUTE


Victor Fox, MD, Division of Gastroenterology, Hepa-ology, and Nutrition, Children’s Hospital Boston, Har-ard Medical School, Boston, Massachusetts

David Katzka, MD, University of Pennsylvania Schoolf Medicine, Division of Gastroenterology, Philadelphia,ennsylvaniaSusan R. Orenstein, MD, Division of Pediatric Gastro-

nterology, University of Pittsburgh Medical School,ittsburgh, PennsylvaniaEric Hassall, MBChB, FRCPC, Division of Gastroenter-

logy, BC Children’s Hospital, University of British Co-umbia, Vancouver, British Columbia, Canada

Allergy

Chair: Marc E. Rothenberg, MD, PhD, Division ofllergy and Immunology, Cincinnati Children’s Hospitaledical Center, University of Cincinnati College of Med-

cine, Cincinnati, OhioJonathan Spergel, MD, Division of Allergy and Immu-

ology, The Children’s Hospital of Philadelphia, Univer-ity of Pennsylvania School of Medicine, Philadelphia,ennsylvaniaAmal Assa’ad, MD, Division of Allergy and Immunol-

gy, Cincinnati Children’s Hospital Medical Center, Uni-ersity of Cincinnati College of Medicine, Cincinnati,hioSeema Aceves, MD, PhD, Allergy Immunology Divi-

ion, Rady Children’s Hospital, San Diego, University ofalifornia, San Diego, San Diego, CaliforniaBarry K. Wershil, MD, Division of Pediatric GI andutrition, Albert Einstein College of Medicine, The Chil-ren’s Hospital at Montefiore, Bronx, New YorkThomas Platts-Mills, MD, PhD, Division of Allergy and

linical Immunology, University of Virginia, Charlottes-ille, VA

Treatment

Chair: Chris Liacouras, MD, Division of Gastro-nterology, Hepatology, and Nutrition, University ofennsylvania School of Medicine, The Children’s Hospi-al of Philadelphia, Philadelphia, Pennsylvania

Chair: Phil E. Putnam, MD, FAAP, Division of Pediat-ic Gastroenterology, Hepatology, and Nutrition, Cincin-ati Children’s Hospital Medical Center, University ofincinnati College of Medicine, Cincinnati, OhioTusar Desai, MD, William Beaumont Hospital, De-

artment of Internal Medicine, Royal Oak Michigan,loomfield Hills, MichiganSeema Khan, MD, Thomas Jefferson University, Phila-

elphia, Pennsylvania; and Division of Gastroenterologynd Nutrition, Alfred I duPont Hospital for Children,ilmington, DelawareB UK Li, MD, Division of Pediatric Gastroenterology,
edical College of Wisconsin, Milwaukee, Wisconsin
Amir F. Kagalwalla, MD, Division of Pediatric Gastro-nterology, University of Illinois, Chicago, Chicago, Illi-oisRichard J. Noel, MD, Division of Pediatric Gastroen-

erology, Medical College of Wisconsin, Milwaukee, Wis-onsin

References

1. Dahms BB. Reflux esophagitis: sequelae and differential diag-nosis in infants and children including eosinophilic esophagitis.Pediatr Dev Pathol 2004;7:5–16.

2. Ngo P, Furuta GT, Antonioli DA, et al. Eosinophils in the esoph-agus—peptic or allergic eosinophilic esophagitis? Case seriesof three patients with esophageal eosinophilia. Am J Gastroen-terol 2006;101:1666–1670.

3. Zimmerman SL, Levine MS, Rubesin SE, et al. Idiopathic eosin-ophilic esophagitis in adults: the ringed esophagus. Radiology2005;236:159–165.

4. Straumann A, Spichtin HP, Grize L, et al. Natural history ofprimary eosinophilic esophagitis: a follow-up of 30 adult pa-tients for up to 11.5 years. Gastroenterology 2003;125:1660–1669.

5. Straumann A, Kristl J, Conus S, et al. Cytokine expression inhealthy and inflamed mucosa: probing the role of eosinophils inthe digestive tract. Inflamm Bowel Dis 2005;11:720–726.

6. Remedios M, Campbell C, Jones DM, et al. Eosinophilic esoph-agitis in adults: clinical, endoscopic, histologic findings, andresponse to treatment with fluticasone propionate. GastrointestEndosc 2006;63:3–12.

7. Potter JW, Saeian K, Staff D, et al. Eosinophilic esophagitis inadults: an emerging problem with unique esophageal features.Gastrointest Endosc 2004;59:355–361.

8. Parfitt JR, Gregor JC, Suskin NG, et al. Eosinophilic esophagitisin adults: distinguishing features from gastroesophageal refluxdisease: a study of 41 patients. Mod Pathol 2006;19:90–96.

9. Kaplan M, Mutlu EA, Jakate S, et al. Endoscopy in eosinophilicesophagitis: “feline” esophagus and perforation risk. Clin Gas-troenterol Hepatol 2003;1:433–437.

10. Desai TK, Stecevic V, Chang CH, et al. Association of eosino-philic inflammation with esophageal food impaction in adults.Gastrointest Endosc 2005;61:795–801.

11. Croese J, Fairley SK, Masson JW, et al. Clinical and endoscopicfeatures of eosinophilic esophagitis in adults. Gastrointest En-dosc 2003;58:516–522.

12. Baxi S, Gupta SK, Swigonski N, et al. Clinical presentation ofpatients with eosinophilic inflammation of the esophagus. Gas-trointest Endosc 2006;64:473–478.

13. Attwood SE, Lewis CJ, Bronder CS, et al. Eosinophilic oesoph-agitis: a novel treatment using Montelukast. Gut 2003;52:181–185.

14. Attwood S, Smyrk T, Demeester T, et al. Esophageal eosino-philia with dysphagia. A distinct clinicopathologic syndrome. DigDis Sci 1993;38:109–116.

15. Arora AS, Perrault J, Smyrk TC. Topical corticosteroid treatmentof dysphagia due to eosinophilic esophagitis in adults. MayoClin Proc 2003;78:830–835.

16. Walsh S, Antonioli D, Goldman H, et al. Allergic esophagitis inchildren—a clinicopathological entity. Am J Surg Pathol 1999;23:390–396.

17. Teitelbaum J, Fox V, Twarog F, et al. Eosinophilic esophagitis inchildren: immunopathological analysis and response to flutica-sone propionate. Gastroenterology 2002;122:1216–1225.

18. Steiner SJ, Gupta SK, Croffie JM, et al. Correlation betweennumber of eosinophils and reflux index on same day esophagealbiopsy and 24 hour esophageal pH monitoring. Am J Gastroen-
terol 2004;99:801–805.

AG

AIN

STIT

UTE


19. Sant’Anna AM, Rolland S, Fournet JC, et al. Eosinophilic esoph-agitis in children: symptoms, histology, and pH probe results.J Pediatr Gastroenterol Nutr 2004;39:373–377.

20. Orenstein SR, Shalaby TM, Di Lorenzo C, et al. The spectrum ofpediatric eosinophilic esophagitis beyond infancy: a clinical se-ries of 30 children. Am J Gastroenterol 2000;95:1422–1430.

21. Noel RJ, Putnam PE, Collins MH, et al. Clinical and immuno-pathologic effects of swallowed fluticasone for eosinophilicesophagitis. Clin Gastroenterol Hepatol 2004;2:568–575.

22. Liacouras CA, Spergel JM, Ruchelli E, et al. Eosinophilic esoph-agitis: a 10-year experience in 381 children. Clin GastroenterolHepatol 2005;3:1198–1206.

23. Kelly KJ, Lazenby AJ, Rowe PC, et al. Eosinophilic esophagitisattributed to gastroesophageal reflux: improvement with anamino acid-based formula. Gastroenterology 1995;109:1503–1512.

24. Gupta SK, Fitzgerald JF, Kondratyuk T, et al. Cytokine expres-sion in normal and inflamed esophageal mucosa: a study intothe pathogenesis of allergic eosinophilic esophagitis. J PediatrGastroenterol Nutr 2006;42:22–26.

25. Fox VL, Nurko S, Teitelbaum JE, et al. High-resolution EUS inchildren with eosinophilic “allergic” esophagitis. GastrointestEndosc 2003;57:30–36.

26. Esposito S, Marinello D, Paracchini R, et al. Long-term follow-upof symptoms and peripheral eosinophil counts in seven childrenwith eosinophilic esophagitis. J Pediatr Gastroenterol Nutr 2004;38:452–456.

27. De Angelis P, Markowitz JE, Torroni F, et al. Paediatric eosino-philic oesophagitis: towards early diagnosis and best treatment.Dig Liver Dis 2006;38:245–251.

28. Cheung KM, Oliver MR, Cameron DJ, et al. Esophageal eosino-philia in children with dysphagia. J Pediatr Gastroenterol Nutr2003;37:498–503.

29. Blanchard C, Wang N, Stringer KF, et al. Eotaxin-3 and a uniquelyconserved gene-expression profile in eosinophilic esophagitis.J Clin Invest 2006;116:536–547.

30. Gupta SK, Peters-Golden M, Fitzgerald JF, et al. Cysteinyl leu-kotriene levels in esophageal mucosal biopsies of children witheosinophilic inflammation: are they all the same? Am J Gastro-enterol 2006;101:1125–1128.

31. Thompson DM, Arora AS, Romero Y, et al. Eosinophilic esoph-agitis: its role in aerodigestive tract disorders. Otolaryngol ClinNorth Am 2006;39:205–221.

32. Noel RJ, Putnam PE, Rothenberg ME. Eosinophilic esophagitis.N Engl J Med 2004;351:940–941.

33. Straumann A, Simon HU. Eosinophilic esophagitis: escalatingepidemiology? J Allergy Clin Immunol 2005;115:418–419.

34. Patel SM, Falchuk KR. Three brothers with dysphagia caused byeosinophilic esophagitis. Gastrointest Endosc 2005;61:165–167.

35. Meyer GW. Eosinophilic esophagitis in a father and a daughter.Gastrointest Endosc 2005;61:932.

36. Straumann A, Rossi L, Simon HU, et al. Fragility of the esoph-ageal mucosa: a pathognomonic endoscopic sign of primaryeosinophilic esophagitis? Gastrointest Endosc 2003;57:407–412.

37. Inns S, Wyeth J, Lamont D, et al. Eosinophilic oesophagitis: anemerging important cause for undiagnosed dysphagia. N Z MedJ 2006;119:U1889.

38. Vakil N, van Zanten SV, Kahrilas P, et al. The Montreal definitionand classification of gastroesophageal reflux disease: a globalevidence-based consensus. Am J Gastroenterol 2006;101:1900–1920.

39. Vasilopoulos S, Murphy P, Auerbach A, et al. The small-caliberesophagus: an unappreciated cause of dysphagia for solids inpatients with eosinophilic esophagitis. Gastrointest Endosc
2002;55:99–106.
40. Sundaram S, Sunku B, Nelson SP, et al. Adherent whiteplaques: an endoscopic finding in eosinophilic esophagitis. J Pe-diatr Gastroenterol Nutr 2004;38:208–212.

41. Straumann A, Spichtin HP, Bucher KA, et al. Eosinophilic esoph-agitis: red on microscopy, white on endoscopy. Digestion 2004;70:109–116.

42. Siafakas CG, Ryan CK, Brown MR, et al. Multiple esophagealrings: an association with eosinophilic esophagitis: case reportand review of the literature. Am J Gastroenterol 2000;95:1572–1575.

43. Nurko S, Teitelbaum JE, Husain K, et al. Association of Schatzkiring with eosinophilic esophagitis in children. J Pediatr Gastro-enterol Nutr 2004;38:436–441.

44. Lucendo Villarin AJ, Carrion Alonso G, Navarro Sanchez M, et al.Eosinophilic esophagitis in adults, an emerging cause of dys-phagia. Description of 9 cases. Rev Esp Enferm Dig 2005;97:229–239.

45. Lim JR, Gupta SK, Croffie JM, et al. White specks in the esoph-ageal mucosa: an endoscopic manifestation of non-reflux eosin-ophilic esophagitis in children. Gastrointest Endosc 2004;59:835–838.

46. Langdon DE. “Congenital” esophageal stenosis, corrugatedringed esophagus, and eosinophilic esophagitis. Am J Gastro-enterol 2000;95:2123–2124.

47. Khan S, Orenstein SR, Di Lorenzo C, et al. Eosinophilic esoph-agitis: strictures, impactions, dysphagia. Dig Dis Sci 2003;48:22–29.

48. Gupta SK, Fitzgerald JF, Chong SK, et al. Vertical lines in distalesophageal mucosa (VLEM): a true endoscopic manifestation ofesophagitis in children? Gastrointest Endosc 1997;45:485–489.

49. Ahmed A, Matsui S, Soetikno R. A novel endoscopic appearance ofidiopathic eosinophilic esophagitis. Endoscopy 2000;32:S33.

50. Morrow JB, Vargo JJ, Goldblum JR, et al. The ringed esophagus:histological features of GERD. Am J Gastroenterol 2001;96:984–989.

51. Straumann A, Bauer M, Fischer B, et al. Idiopathic eosinophilicesophagitis is associated with a T(H)2-type allergic inflamma-tory response. J Allergy Clin Immunol 2001;108:954–961.

52. Steiner SJ, Kernek KM, Fitzgerald JF. Severity of basal cellhyperplasia differs in reflux versus eosinophilic esophagitis.J Pediatr Gastroenterol Nutr 2006;42:506–509.

53. Spergel JM, Brown-Whitehorn T. The use of patch testing in thediagnosis of food allergy. Curr Allergy Asthma Rep 2005;5:86–90.

54. Ruchelli E, Wenner W, Voytek T, et al. Severity of esophagealeosinophilia predicts response to conventional gastroesopha-geal reflux therapy. Pediatr Develop Pathol 1999;2:15–18.

55. Markowitz JE, Spergel JM, Ruchelli E, et al. Elemental diet is aneffective treatment for eosinophilic esophagitis in children andadolescents. Am J Gastroenterol 2003;98:777–782.

56. Liacouras C, Wenner W, Brown K, et al. Primary eosinophilicesophagitis in children: successful treatment with oral cortico-steroids. J Pediatr Gastroenterol Nutr 1998;26:380–385.

57. Gonsalves N, Policarpio-Nicolas M, Zhang Q, et al. Histopatho-logic variability and endoscopic correlates in adults with eosin-ophilic esophagitis. Gastrointest Endosc 2006;64:313–319.

58. Vitellas KM, Bennett WF, Bova JG, et al. Idiopathic eosinophilicesophagitis. Radiology 1993;186:789–793.

59. Feczko PJ, Halpert RD, Zonca M. Radiographic abnormalities ineosinophilic esophagitis. Gastrointest Radiol 1985;10:321–324.

60. Dobbins JW, Sheahan DG, Behar J. Eosinophilic gastroenteritiswith esophageal involvement. Gastroenterology 1977;72:1312–1316.

61. Picus D, Frank PH. Eosinophilic esophagitis. AJR Am J Roentge-
nol 1981;136:1001–1003.

1

1

1

1

1

1

1

1

AG

AIN

STITUTE


62. Munch R, Kuhlmann U, Makek M, et al. Eosinophilic esophagi-tis, a rare manifestation of eosinophilic gastroenteritis. SchweizMed Wochenschr 1982;112:731–734.

63. Matzinger MA, Daneman A. Esophageal involvement in eosino-philic gastroenteritis. Pediatr Radiol 1983;13:35–38.

64. Landres RT, Kuster GG, Strum WB. Eosinophilic esophagitis in apatient with vigorous achalasia. Gastroenterology 1978;74:1298–1301.

65. Winter HS, Madara JL, Stafford RJ, et al. Intraepithelial eosino-phils: a new diagnostic criterion for reflux esophagitis. Gastro-enterology 1982;83:818–823.

66. Shub MD, Ulshen MH, Hargrove CB, et al. Esophagitis: a fre-quent consequence of gastroesophageal reflux in infancy. J Pe-diatr 1985;107:881–884.

67. Leape LL, Bhan I, Ramenofsky ML. Esophageal biopsy in thediagnosis of reflux esophagitis. J Pediatr Surg 1981;16:379–384.

68. Hyams JS, Ricci A Jr, Leichtner AM. Clinical and laboratorycorrelates of esophagitis in young children. J Pediatr Gastroen-terol Nutr 1988;7:52–56.

69. Brown LF, Goldman H, Antonioli DA. Intraepithelial eosinophilsin endoscopic biopsies of adults with reflux esophagitis. Am JSurg Pathol 1984;8:899–905.

70. Lee RG. Marked eosinophilia in esophageal mucosal biopsies.Am J Surg Pathol 1985;9:475–479.

71. Vitellas KM, Bennett WF, Bova JG, et al. Radiographic manifes-tations of eosinophilic gastroenteritis. Abdom Imaging 1995;20:406–413.

72. Faubion WA Jr, Perrault J, Burgart LJ, et al. Treatment of eosin-ophilic esophagitis with inhaled corticosteroids. J Pediatr Gas-troenterol Nutr 1998;27:90–93.

73. Spergel JM, Andrews T, Brown-Whitehorn TF, et al. Treatment ofeosinophilic esophagitis with specific food elimination diet di-rected by a combination of skin prick and patch tests. AnnAllergy Asthma Immunol 2005;95:336–343.

74. Kato M, Kephart GM, Morikawa A, et al. Eosinophil infiltrationand degranulation in normal human tissues: evidence for eosin-ophil degranulation in normal gastrointestinal tract. Int ArchAllergy Immunol 2001;125(Suppl 1):55–58.

75. DeBrosse CW, Case JW, Putnam PE, et al. Quantity and distri-bution of eosinophils in the gastrointestinal tract of children.Pediatr Dev Pathol 2006;9:210–218.

76. Mishra A, Rothenberg ME. Intratracheal IL-13 induces eosino-philic esophagitis by an IL-5, eotaxin-1, and STAT6-dependentmechanism. Gastroenterology 2003;125:1419–1427.

77. Mishra A, Hogan SP, Brandt EB, et al. IL-5 promotes eosinophiltrafficking to the esophagus. J Immunol 2002;168:2464–2469.

78. Akei HS, Mishra A, Blanchard C, et al. Epicutaneous antigenexposure primes for experimental eosinophilic esophagitis inmice. Gastroenterology 2005;129:985–994.

79. Mishra A, Hogan SP, Brandt EB, et al. An etiological role foraeroallergens and eosinophils in experimental esophagitis. J ClinInvest 2001;107:83–90.

80. Onbasi K, Sin AZ, Doganavsargil B, et al. Eosinophil infiltrationof the oesophageal mucosa in patients with pollen allergy duringthe season. Clin Exp Allergy 2005;35:1423–1431.

81. Fogg MI, Ruchelli E, Spergel JM. Pollen and eosinophilic esoph-agitis. J Allergy Clin Immunol 2003;112:796–797.

82. Konikoff MR, Blanchard C, Kirby C, et al. Potential of bloodeosinophils, eosinophil-derived neurotoxin, and eotaxin-3 as bi-omarkers of eosinophilic esophagitis. Clin Gastroenterol Hepa-tol 2006;4:1328–1336.

83. Simon D, Marti H, Heer P, et al. Eosinophilic esophagitis isfrequently associated with IgE-mediated allergic airway dis-eases. J Allergy Clin Immunol 2005;115:1090–1092.

84. Bax KC, Gupta SK. Allergic eosinophilic esophagitis. Indian
J Pediatr 2006;73:919–925.
85. Kagalwalla AF, Sentongo TA, Ritz S, et al. Effect of six-food elimi-nation diet on clinical and histologic outcomes in eosinophilicesophagitis. Clin Gastroenterol Hepatol 2006;4:1097–1102.

86. Rostenberg A, Sulzberger M. Some results of patch tests. ArchDermatol 1937;35:433.

87. Mitchell EB, Crow J, Chapman SS, et al. Basophils in allergen-inducedpatch test sites in atopic dermatitis. Lancet 1982;1:127.

88. Darsow U, Ring J. Airborne and dietary allergens in atopic ec-zema: a comprehensive review of diagnostic tests. J Exp Der-matol 2000;25:544.

89. Leung D, Bieber T. Atopic dermatitis. Lancet 2003;361:151.90. Kalach N, Soulaines P, de Boissieu D, et al. A pilot study of the

usefulness and safety of a ready-to-use atopy patch test (Dialler-test) versus a comparator (Finn Chamber) during cow’s milk allergyin children. J Allergy Clin Immunol 2005;116:1321–1326.

91. Isolauri E, Turjanmaa K. Combined skin prick and patch testingenhances identification of food allergy in infants with atopicdermatitis. J Allergy Clin Immunol 1996;97:9–15.

92. Isolauri E, Pelto L, Nuutila J, et al. Altered expression of IgG andcomplement receptors indicates a significant role of phagocytesin atopic dermatitis. J Allergy Clin Immunol 1997;99:707–713.

93. Spergel JM, Beausoleil JL, Mascarenhas M, et al. The use of skinprick tests and patch tests to identify causative foods in eosino-philic esophagitis. J Allergy Clin Immunol 2002;109:363–368.

94. Rance F. What is the optimal occlusion time for the atopy patchtest in the diagnosis of food allergies in children with atopicdermatitis? Pediatr Allergy Immunol 2004;15:93–96.

95. Untersmayr E, Scholl I, Swoboda I, et al. Antacid medicationinhibits digestion of dietary proteins and causes food allergy: afish allergy model in BALB/c mice. J Allergy Clin Immunol 2003;112:616–623.

96. Untersmayr E, Bakos N, Scholl I, et al. Anti-ulcer drugs promoteIgE formation toward dietary antigens in adult patients. FASEB J2005;19:656–658.

97. Cantu P, Velio P, Prada A, et al. Ringed oesophagus and idio-pathic eosinophilic oesophagitis in adults: an association in twocases. Dig Liver Dis 2005;37:129–134.

98. Langdon DE. Response to Straumann et al: primary eosinophilicesophagitis. Gastroenterology 2004;127:364–365.

99. Aceves SS, Bastian JF, Newbury RO, Dohil R. Oral viscousbudesonide: A potential new therapy for eosinophilic esophagi-tis in children. Am J Gastroenterol 2007;102:1–9.

00. Lipworth BJ. Systemic adverse effects of inhaled corticosteroidtherapy: a systematic review and meta-analysis. Arch InternMed 1999;159:941–955.

01. Simon D, Braathen LR, Simon HU. Anti-interleukin-5 antibodytherapy in eosinophilic diseases. Pathobiology 2005;72:287–292.

02. Garrett JK, Jameson SC, Thomson B, et al. Anti-interleukin-5(mepolizumab) therapy for hypereosinophilic syndromes. J Al-lergy Clin Immunol 2004;113:115–119.

03. Dellon ES, Aderoju A, Woosley JT, et al. Variability in diagnosticcriteria for eosinophilic esophagitis: a systematic review. Am JGastroenterology 2007 [in press].

04. Byrne KR, Panagiotakis PH, Hilden K, et al. Retrospective anal-ysis of esophageal food impaction: differences in etiology byage and gender. Dig Dis Sci 2007;52:717–721.

05. Lee GS, Craig PI, Freiman JS, et al. Intermittent dysphagia forsolids associated with a multiringed esophagus: clinical fea-tures and response to dilatation. Dysphagia 2007;22:55–62.

06. Zink DA, Amin M, Gebara S, et al. Familial dysphagia andeosinophilia. Gastrointest Endosc 2007;65:330–334.

07. Quaglietta L, Coccorullo P, Miele E, et al. Eosinophilic oesoph-agitis and coeliac disease: is there an association? Aliment
Pharmacol Ther 2007;26:487–493.

1

1

1

1

1

1

1

1

1

1

1

1

1

1

Qc

1

NcAsgt

Mg

TSbCTesEPdtNNJnnsCeSpKBeP


08. Ogden S, Denyer ME, Wilkinson SM. Erythema nodosum andeosinophilic oesophagitis: more than a chance association?Br J Dermatol 2007;156:1388–1389.

09. Balatsinou C, Milano A, Caldarella MP, et al. Eosinophilicesophagitis is a component of the anticonvulsant hypersensi-tivity syndrome: description of two cases. Dig Liver Dis 2007 [inpress].

10. Assa’ad AH, Putnam PE, Collins MH, et al. Pediatric patientswith eosinophilic esophagitis: an 8-year follow-up. J Allergy ClinImmunol 2007;119:731–738.

11. Ronkainen J, Talley NJ, Aro P, et al. Prevalence of oesophagealeosinophils and eosinophilic oesophagitis in adults: the popu-lation-based Kalixanda study. Gut 2007;56:615–620.

12. Vanderheyden AD, Petras RE, DeYoung BR, et al. Emerging eosin-ophilic (allergic) esophagitis: increased incidence or increasedrecognition? Arch Pathol Lab Med 2007;131:777–779.

13. Yamazaki K, Murray JA, Arora AS, et al. Allergen-specific in vitrocytokine production in adult patients with eosinophilic esoph-agitis. Dig Dis Sci 2006;51:1934–1941.

14. Stein ML, Collins MH, Villanueva JM, et al. Anti-IL-5 (mepoli-zumab) therapy for eosinophilic esophagitis. J Allergy Clin Im-munol 2006;118:1312–1319.

15. Spergel JM, Brown-Whitehorn T, Beausoleil JL, et al. Predictivevalues for skin prick test and atopy patch test for eosinophilicesophagitis. J Allergy Clin Immunol 2007;119:509–511.

16. Konikoff MR, Noel RJ, Blanchard C, et al. A randomized, double-blind, placebo-controlled trial of fluticasone propionate for pe-diatric eosinophilic esophagitis. Gastroenterology 2006;131:1381–1391.

17. Mishra A, Schlotman J, Wang M, et al. Critical role for adaptiveT-cell immunity in experimental eosinophilic esophagitis inmice. J Leukoc Biol 2007;81:916–924.

18. Aceves SS, Newbury RO, Dohil R, et al. Esophageal remodelingin pediatric eosinophilic esophagitis. J Allergy Clin Immunol2007;119:206–212.

19. Lucendo AJ, Navarro M, Comas C, et al. Immunophenotypiccharacterization and quantification of the epithelial inflamma-tory infiltrate in eosinophilic esophagitis through stereology: ananalysis of the cellular mechanisms of the disease and theimmunologic capacity of the esophagus. Am J Surg Pathol2007;31:598–606.

20. Kirsch R, Bokhary R, Marcon MA, et al. Activated mucosal mastcells differentiate eosinophilic (allergic) esophagitis from gas-troesophageal reflux disease. J Pediatr Gastroenterol Nutr
2007;44:20–26. n
21. Bullock JZ, Villanueva JM, Blanchard C, et al. Interplay of adaptiveth2 immunity with eotaxin-3/c-C chemokine receptor 3 in eosino-philic esophagitis. J Pediatr Gastroenterol Nutr 2007;45:22–31.

Address requests for reprints to: Chair of the Clinical Practice anduality Management Committee, American Gastroenterological Asso-iation (AGA) Institute, 4930 Del Ray Ave, Bethesda, Maryland 20814.Dr Furuta’s current address is: The Children’s Hospital, 13123 East

6th Avenue, Aurora, Colorado 80045. fax: (720) 777-8025.Supported by a small conference grant R13 DK076672 from the

ational Institutes of Health to The FIGER Symposium; North Ameri-an Society of Pediatric Gastroenterology, Hepatology; the Americancademy of Allergy, Asthma, and Immunology; The American Partner-hip for Eosinophilic Diseases; a philanthropic contribution from arateful family; and educational grants (AstraZeneca, Abbot Labora-ories, Nutricia, TAP, Ception, GlaxoSmithKline).

The authors thank Sheila Crowe, MD, for insights and support andargaret Stallings, Sandy Fasold, and Kim Rose for the efficient and

racious administrative support.Conflict of interest disclosures: Glenn T. Furuta, consultant, Ception

herapeutics; speaker’s bureau, TAP; Chris A. Liacouras, consultant,HS, Nutricia, Ception, Ross; grant/research support, Wyeth; speaker’sureau, TAP, Merck; Margaret H. Collins, consultant, GlaxoSmithKline,eption Therapeutics; Sandeep K. Gupta, consultant, GlaxoSmithKline,AP, AstraZeneca, Salix; speaker’s bureau, Ross Products, TAP, AstraZen-ca; educational grant, Ross Products; Christopher Justinich, no disclo-ures; Phil E. Putnam, no disclosures; Peter A Bonis, no disclosures;ric Hassall, consultant, TAP Pharmaceuticals, Abbott Canada, Altanaharma; clinical research grant, AstraZeneca; Alex Straumann, noisclosures; Marc E. Rothenberg, consultant, Merck, Ception Therapeu-ics, GlaxoSmithKline, MedaCorp; speaker’s bureau, Merck; Samuelurko, grant/research support, Wyeth pharmaceutica, TAP, Sucampo;irmala Gonsalves, consultant, Medacorp, Ception Therapeutics;onathan Markowitz, consultant, Ception Therapeutics; Don Antonioli,o disclosures; Eduardo Ruchelli, no disclosures; Hector Melin-Aldana,o disclosures; Margret Magid, no disclosures; Ikuo Hirano, no disclo-ures; David Katzka, no disclosures; Susan R. Orenstein, consultant,eption Therapeutics, TAP, Braintree, AstraZeneca, Wyeth, Bristol My-rs Squibb, McNeil; grant/research support, Braintree; Jonathan M.pergel, consultant, Novartis, GlaxoSmithKline; grant/research sup-ort, Novartis, Nutricia; speaker’s bureau, AstraZeneca, GlaxoSmith-line; Amal Assa’ad, no disclosures; Seema Aceves, no disclosures;arry K. Wershil, consultant, AP Pharmaceuticals, AstraZeneca; speak-r’s bureau, Shire; educational grant, TAP Pharmaceuticals; Thomaslatts-Mills, no disclosures; Tusar Desai, no disclosures; Seema Khan,
o disclosures; B Li, no disclosures; Amir F. Kagalwalla, no disclosures.
AG

AIN

STIT

UTE

aga institute - naspghan

Documents