Densities of rectal peptide YY and somatostatin cells as biomarkers for the diagnosis of irritable bowel syndrome

(1)

ContentslistsavailableatScienceDirect

Peptides

jo u r n al h om ep age :w w w . e l s e v i e r . c o m / l o c a t e / p e p t i d e s

Densities of rectal peptide YY and somatostatin cells as biomarkers for the diagnosis of irritable bowel syndrome

Magdy El-Salhy

^a,b,c,∗

, Jan Gunnar Hatlebakk

^b,c

, Odd Helge Gilja

^b,c

, Trygve Hausken

^b,c

aDivisionofGastroenterology,DepartmentofMedicine,StordHospital,Stord,Norway

bDivisionofGastroenterology,DepartmentofClinicalMedicine,UniversityofBergen,Bergen,Norway

cNationalCentreforFunctionalGastrointestinalDisorders,DepartmentofMedicine,HaukelandUniversityHospital,Bergen,Norway

a r t i c l e i n f o

Articlehistory:

Received24February2015 Accepted27February2015 Availableonline9March2015

Keywords:

Biomarkers Diagnosis

Irritablebowelsyndrome Immunohistochemistry PYY

Somatostatin

a b s t r a c t

Irritablebowelsyndrome(IBS)isacommonchronicdisorder.IBSdiagnosisisadiagnosisofexclusion sincetherearenobloodtests,radiologicalorendoscopicexaminationsforthisdisorder.Althoughseveral attemptshavebeenmadetodevelopasymptoms-baseddiagnosis,suchsystemsarenotwidelyusedin clinics.SeveraltestsandexaminationsmeasuringpathologicalﬁndingsinIBShavebeenconsidered forthediagnosisofIBS,butnoneofthemhasprovedusefulasabiomarker.Abnormalitiesinthecell densitiesofrectalpeptideYY(PYY)andsomatostatincellshavebeenreportedinIBSpatients.Theaim ofthepresentstudywastodeterminetheutilityoftheseabnormalitiesasbiomarkersforthediagnosis ofIBS.PatientswithIBSestablishedaccordingtoRomeIIIcriteria(n=101)wereincludedinthisstudy (71femalesand30maleswithameanageof35years;range18–61years),and62healthysubjects (38femalesand24maleswithameanageof41years;range18–65years)wererecruitedascontrols.

Boththepatientsandcontrolsunderwentcolonoscopyduringwhichrectalbiopsysamplesweretaken.

ThetissuesampleswereimmunostainedforPYYandsomatostatin,andthenumberofstainedcells wasquantifiedrelativetoboththeareaofepithelialcellsandpermicroscopicfield.ThedensityofPYY cellswassignificantlylowerinIBSpatientsthaninthehealthycontrols(P<0.0001);receiveroperator characteristic(ROC)analysisrevealedanareaundertheROCcurve(AUC)of0.99.Thesomatostatincell densityinIBSpatientswashigherthaninthecontrols(P<0.0001);ROCanalysisrevealedanAUCof0.86.

ThedensitiesoftherectalPYYandsomatostatincellsappeartobeclinicallyeffectivebiomarkersforIBS.

Furthermore,measurementoftheseparametersisinexpensive,rapidanddoesnotrequireconsiderable experienceorsophisticatedequipment.

Introduction

Irritablebowelsyndrome(IBS)isacommonchronicdisorder thataffectsbetween5%and20%ofthepopulationoftheWestern world[1,2],andyettherearecurrentlynobloodtests,radiological orendoscopicexaminationsfordiagnosingthedisorder.Although severalattemptshavebeenmadetoobtainadiagnosisbasedon symptomassessments similar to those used in psychiatry and rheumatology[3–9],IBSdiagnosisinclinicalpracticeremains a processofexclusion[1,10–12].Thus,adiagnosisofIBScanonly

∗Correspondingauthorat:SectionforGastroenterology,DepartmentofMedicine, StordHospital,Box4000,5409Stord,Norway.Tel.:+4753491000;

fax:+475349100.

E-mailaddresses:[email protected](M.El-Salhy),

[email protected](J.G.Hatlebakk), [email protected] (O.H.Gilja),[email protected](T.Hausken).

bereachedbyconductingextensiveexpensiveexaminationsand teststoexcludeorganiccausesforthepatient’ssymptoms.

Severalabnormalities in the gastrointestinalendocrine cells have been described in IBSpatients [13–27].These abnormali- tiesarebelievedtoplayanimportantroleinthepathophysiology of IBS and represent a potentialtool for the treatment of this disorder[28,29].Apreviousstudy[30]foundthecelldensityof rectalpeptideYY(PYY)cellstobedecreased,andthatofsomato- statinto be increased relative to healthy subjects.It hasbeen suggestedthattheseabnormalitiescanbeusedinthediagnosis ofIBS[30].However,thatstudyincluded50patientsand27con- trols,andthesubtypesofIBSoftheincludedpatientswereonly those withdiarrhea or constipation as thepredominant symptom.Thepresentstudywasundertakentoinvestigatetheefﬁcacy ofthedensitiesofrectalPYYandsomatostatincellsasbiomark- ersforthediagnosisofIBSinalargercohortofIBSpatientsand withalargergroupofhealthycontrolsforcomparisonthanwas usedin thepreviousstudy.Furthermore,IBS patientswithany http://dx.doi.org/10.1016/j.peptides.2015.02.008

(2)

oftheIBSsubtypes—constipation-predominant(IBS-C),diarrhea- predominant(IBS-D)ormixed(IBS-M)—wereincluded.

Materialandmethods

Patientsandcontrols

OnehundredandonepatientssufferingfromIBSaccordingto RomeIIIcriteria[9]wereincludedinthisstudy.Pregnantorlactat- ingwomen,patientswithclinicallysignificantsystemicdiseases, drugabuse,seriouspsychiatricdiseases,andabdominalsurgery, withtheexceptionofappendectomy,cesareansectionandhys- terectomywereexcluded.Thecohortcomprised71femalesand30 maleswithameanageof35years(range18–61years);45patients hadIBS-D,14hadIBS-Mand42hadIBS-C.Allofthepatientshada longdurationofIBSsymptomsi.e.greaterthan10years,andtheir onsetofIBSsymptomswasnotassociatedwithaboutofgastroen- teritis.Thepatientssubmittedtoaphysicalexaminationandblood teststoexcludeinflammation, liverand pancreasdiseases, and thyroiddysfunction.Celiacdiseasewasexcludedbyhistopatho- logicalexaminationofduodenalbiopsysamples.Forty-fiveofthese patientsusedproton-pumpinhibitorsondemand,18weremed- icatedwithangiotensinII-receptorantagonistagainsthypertonia, 15withstatinsagainsthighbloodcholesterollevels,and2with levothyroxineagainsthypothyroidosis.

The controls comprised 62 subjects (38 females and 24 males; mean age 41 years, range 18–65 years) who under- went colonoscopywithrectal biopsysampling.The reasonsfor colonoscopy in these subjects were gastrointestinal bleeding, wherethebleedingsourcewasidentiﬁedashemorrhoids(n=36), or angiodysplasia (n=4), and otherwise healthy subjects with healthworriescausedbyarelative(s)beingdiagnosedwithcolon carcinoma(n=22).Fiveofthesesubjectsweretreatedforhyper- toniawithangiotensinII-receptorantagonist,4withstatinsand2 withlevothyroxine.

ThestudywasperformedinaccordancewiththeDeclarationof HelsinkiandwasapprovedbytheRegionalCommitteeforMedi- calandHealthResearchEthicsWest,Bergen,Norway.Allsubjects providedverbalandwrittenconsenttoparticipate.

Colonoscopy,histopathologyandimmunohistochemistry

A standard colonoscopy procedurewas performed onall of thepatientsand controls, and four biopsysamples weretaken fromthedorsal wallofthe rectumabout 15cm fromtheanus ineach patient.Thebiopsysampleswereﬁxedovernightin4%

bufferedparaformaldehyde,embeddedinparafﬁnandthensec- tionedat5␮m.Thesectionswerestainedwithhematoxylin–eosin andimmunostainedwithultraViewuniversalDABdetectionkit (v1.02.0018, Venata Medical Systems, Basel,Switzerland) using theBenchMark Ultra IHC/ISHstaining module (Venata Medical Systems).Thesectionswereincubatedwiththeprimaryantibod- iesfor 32minat37^◦C. Theprimaryantibodies werepolyclonal anti-porcinepeptideYY(PYY;codeno.PYY11A,AlphaDiagnostic, SanAntonio,TX,USA)andpolyclonalrabbitanti-synthetichuman somatostatin (codeno.A566, Dako, Glostrup, Denmark). Nega- tiveandpositivecontrolswereperformedasdescribedelsewhere [31,32].

Quantiﬁcationofimmunoreactivecelldensities

ThedensitiesofrectalPYYandsomatostatincellswerequan- tiﬁedbycomputerizedimageanalysisandcountingthenumber ofpositivecellsinamicroscopicﬁeld.Measurementsusingcom- puterizedimageanalysiswereperformedona computerlinked toamicroscope(BX43,Olympus,Tokyo,Japan)equippedwitha

digitalcamera(DP26,Olympus),usingOlympuscellSensimaging (version1.7) software.The numbersof PYY- and somatostatin- immunoreactive cells were counted in ten randomly chosen microscopicﬁelds,andtheareaofepithelialcellswasmeasured.

Thenumberofendocrinecellsineachﬁeldwascountedmanu- allybypointingandclickingthecomputermouse,andtheareaof theepitheliumcontainingthesecellswasdrawnmanuallyusing thecomputermouse.A×40objectivewasused,forwhicheach frame(ﬁeld)onthemonitorrepresentedatissueareaof0.035mm². Immunostainedsectionsfromthepatientsandcontrolswerecoded and mixed,and measurementsweremadeby thesameperson (M.E.)whowasblindtotheidentity ofthesections(i.e.,IBSor control).

Statisticalanalysis

Differences in gender between patients and controls were testedusingFisher’sexacttest.Differencesintheageproﬁlewere testedusingtheMann–Whitneynon-parametrictest.Differences betweencontrolsandallIBSpatients(IBS-total),andbetweenIBS- D,IBS-MandIBS-CpatientsweretestedusingtheKruskal–Wallis non-parametrictestwithDunn’spost-test.Thedataarepresented asmean±SEMvalues,anddifferenceswithP<0.05wereconsid- eredtobestatisticallysigniﬁcant.

Results

Patientsandcontrols

The genderand age distributions didnot differsigniﬁcantly betweenthepatientsandthecontrols(P=0.2and0.7,respectively).

Colonoscopy,histopathologyandimmunohistochemistry

Theendoscopicandhistopathologicalappearanceofthecolon andrectum wasnormalin boththepatientsandcontrols.PYY- andsomatostatin-immunoreactivecellswerefoundmostlyinthe cryptsofLieberkühninboththepatientsandthecontrols,andwere basket-orﬂask-shaped,sometimeswithalongbasalcytoplasmic process.

Quantiﬁcationofimmunoreactivecelldensity PYYcelldensity

ThedensitiesofPYYcellsasdeterminedbycomputerizedimage analysis and as counted per microscopic field are summarized inTables1and 2anddepictedgraphicallyinFigs.1and2.The Kruskal–Wallistestrevealedsignificantdifferencesbetweenthe IBSpatientsandthecontrols(P<0.0001).Dunn’spost-testrevealed thatPYYcelldensitywaslowerinIBS-totalandallthreeoftheIBS subgroupsrelativetothecontrols(P<0.0001forall).Theresults of receiver operator characteristic(ROC) analysiswith thearea under theROCcurve (AUC),95%confidence intervals (95%CIs) andPvaluesforPYYcelldensityareshowninFigs.3and4.The sensitivityandspecificityforcut-offvaluesof<188cells/mm²and

<6cells/microscopicﬁeldforadiagnosisofIBSaregiveninTable3.

Somatostatincelldensity

Thesomatostatincell densitywassigniﬁcantlyhigher in IBS patients than in the controls (Tables 1 and 2; Figs. 5 and 6;

Kruskal–Wallistest,P<0.0001).Dunn’spost-testrevealedthatthe increased somatostatin cell density was statistically signiﬁcant inIBS-total andalloftheIBS-subgroupsrelative tothecontrols (P<0.0001 forall). ROC curves (Figs. 7and 8)for somatostatin celldensity,andthesensitivityandspeciﬁcityforcut-offvalues

(3)

Table1

DensitiesofPYYandsomatostatincellsincontrolsandpatients(measuredbycomputerizedimageanalysis),expressedasthenumberofcellsrelativetotheareaofepithelial cells.Dataaremean±SEMvalues.

Endocrinecelltype Controls IBS-total IBS-D IBS-M IBS-C

Cells/mm² Cells/mm² Cells/mm² Cells/mm² Cells/mm²

PYY 430±9 144±4^*** 145±6^*** 128±6^*** 148±5^***

Somatostatin 156±7 251±8^*** 265±2^*** 257±1^*** 234±7^***

***P<0.0001vs.controls.

Table2

DensitiesofPYYandsomatostatincellsincontrolsandIBSpatientsexpressedasthenumberofcellspermicroscopicﬁeld.Dataaremean±SEMvalue.

Endocrinecelltype Controls IBS-total IBS-D IBS-M IBS-C

Cells/field Cells/field Cells/field Cells/field Cells/field

PYY 8.4±0.3 4.5±0.1^*** 4.6±0.2^*** 4.5±0.3^*** 4.6±0.2^***

Somatostatin 5.0±0.2 9.8±0.8^*** 8.9±0.2^*** 8.3±0.3^*** 8.4±0.2^***

***P<0.0001vs.controls.

Fig.1. RectalPYY-immunoreactivecellsin(A)acontrolsubjectand(B)apatientwithIBS.

A

Con trols

IBS-total IBS-D

IBS -M

IBS-C 0

200 400 600

* * * *

Number of cells/mm2 epithelium (mean + SE)

B

Controls IBS-total

IBS-D IBS-M

IBS-C 0

2 4 6 8 10

Number of cells/field (mean + SE)

* * * *

Fig.2. DensityofPYYcellsexpressedrelativetotheareaofepithelialcells(A)andasthenumberofcellspermicroscopicﬁeld(B)in10randomlychosenﬁeldsintissue samplestakenfromthecontrolsandIBS-total,IBS-D,IBS-M,andIBS-Cpatients.^***P<0.0001vs.controlgroup.

Table3

SensitivityandspeciﬁcityofthePYYcelldensityasabiomarkerforIBSatcut-offvaluesof<188cells/mm²epitheliumand<6cells/microscopicﬁeld.

IBS-total IBS-D IBS-M IBS-C

Cells/mm² Cells/field Cells/mm² Cells/field Cells/mm² Cells/field Cells/mm² Cells/field

Sensitivity(%) 89 90 87 87 100 100 88 88

Speciﬁcity(%) 87 95 87 89 87 89 87 89

(4)

ROC curve A

0 20 40 60 80 100

0 50 100 150

100% - Specificity%

100%-Sensitivity %

ROC c urve C

0 20 40 60 80 100

0 50 100 150

100%- Sensitivity

ROC curve D

0 20 40 60 80 100

0 50 100 150

ROC curve B

0 20 40 60 80 100

0 50 100 150

Fig.3.ResultsofROCanalysisforPYYcelldensityasmeasuredrelativetotheareaofepithelialcellsin(A)IBS-total(AUC=0.95,95%CI=0.92–0.98,P<0.0001),(B)IBS-D (AUC=0.94,95%CI=0.90–0.98,P<0.0001),(C)IBS-M(AUC=0.99,95%CI=0.97–1.00,P<0.0001)and(D)IBS-C(AUC=0.95,95%CI=0.91–0.98,P<0.0001).AUC=1represents aperfecttest,andAUC=0.5representsanon-discriminatingtest.

ROC curve A

0 20 40 60 80 100

0 50 100 150

ROC curve B

0 20 40 60 80 100

0 50 100 150

ROC c urve C

0 20 40 60 80 100

0 50 100 150

ROC curve D

0 20 40 60 80 100

0 50 100 150

Sensitivity%

Fig.4. ResultsofROCanalysisforPYYcelldensitymeasuredin10randomlychosenmicroscopicﬁeldsin(A)IBS-total(AUC=0.99,95%CI=0.98–1.00,P<0.0001),(B)IBS-D (AUC=0.99,95%CI=0.97–1.00,P<0.0001),(C)IBS–M(AUC=1.0,95%CI=1.00–1.00,P<0.0001)and(D)IBS-C(AUC=0.99,95%CI=0.98–1.00,P<0.0001).

(5)

Fig.5.Somatostatin-immunoreactivecellsinrectaltissuefromacontrolsubject(A)andapatientwithIBS(B).

A

Controls IBS-tota

l IBS

-D IBS

-M IBS-C 0

100 200 300

Number of cells/mm2 epithelium (mean + SE)

*** *** ***

***

B

Controls IBS-total

IBS-D IBS-M

IBS-C 0

5 10 15

Number of cells/field (mean + SE)

* * * *

Fig.6.Somatostatincelldensityexpressedrelativetotheareaofepithelialcells(A)and(B)asthenumberofcellspermicroscopicﬁeldinIBS-total,IBS-D,IBS-MandIBS-C patients.^***P<0.0001vs.controlgroup.

Table4

SensitivityandspeciﬁcityofthesomatostatincelldensityasabiomarkerforIBSatcut-offvaluesof>182cells/mm²epitheliumand>6cells/microscopicﬁeld.

IBS-total IBS-D IBS-M IBS-C

Cells/mm² Cells/field Cells/mm² Cells/field Cells/mm² Cells/field Cells/mm² Cells/field

Sensitivity(%) 91 97 87 96 100 93 93 92

Speciﬁcity(%) 81 90 81 90 81 90 81 90

of>182cells/mm²and>6cells/microscopicﬁeldforadiagnosisof IBSaregiveninTable4.

Discussion

Thediagnosisof IBSisperformedviaaprocessof exclusion, whereby extensive examinations and tests are conducted to exclude other organic diseases [1,10–12]. Attempts have been madetoachieve apositive symptom-baseddiagnosissimilarto thoseusedinpsychiatry[3–9,33],whichculminatedintheestab- lishmentofthesymptoms-basedRomeIdiagnosticcriteriaforIBS in1988byaninternationalgroupofgastroenterologists.Thesecri- teriaweresucceededbyreﬁnementsin1999(RomeIIcriteria)and 2006(RomeIIIcriteria).TheRomecriteriaarenotwidelyusedin currentclinicalpractice[9,11,34]becauseofthepotentialtomiss organicdiseases thatmimic IBSsymptoms buthave adifferent pathophysiologyandtreatments[10,11,34,35].Severaltestsand examinations measuring gut motility, visceral hypersensitivity,

autonomicreactivity,mucosalinﬂammation,fecalproteases,gut ﬂora,serumantibodies,geneexpressionandfoodallergieshave beenconsidered for the diagnosisof IBS, but nonehas proven usefulasabiomarkerforIBSdiagnosis[10–13,36,37].

Apaneloftenserologicalbiomarkersthatarecommonlycon- sideredinthedifferentialdiagnosisforIBSpatients,suchasthose forinﬂammatoryboweldiseaseandcoeliacdisease,wereexam- inedbyLemboetal.asbiomarkersforthediagnosisofIBS[38].

TheseauthorsusedthesmartDiagnosticAlgorithmandfoundan AUCof0.763forthispanel.Jonesetal.addedafurther24sero- logicalbiomarkerstothatpanel(thustotalling34),andexamined themtogetherwith4psychologicalmarkers[39].Thecombination ofthese34markershasbeenreportedtobegoodfordifferenti- atingIBSfromhealth(AUC=0.81),andwasimprovedbyadding thefourpsychologicalmarkers(AUC=0.92)[39].However,theuse ofserologicalbiomarkerscommontoothergastrointestinaldis- easescannotbeconsideredasanIBS-speciﬁcbiomarker;rather,it isyetanothermethodofexclusiondiagnosis.Furthermore,com- biningtheseserologicalmarkerswithpsychologicalmarkers,with

(6)

A

0 50 100

150 0

50 100 150

B

0 50 100

150 0

20 40 60 80 100

C

0 20 40 60 80

0 50 100 150

D

0 20 40 60 80

0 20 40 60 80 100

Fig.7. ROCcurvesforsomatostatincelldensityexpressedrelativetotheareaofepithelialcellsin(A)IBS-total(AUC=0.86,95%CI=0.79–0.92,P<0.0001),(B)IBS-D(AUC=0.83, 95%CI=0.73–0.93,P<0.0001),(C)IBS-M(AUC=0.90,95%CI=0.83–0.97,P<0.0001)and(D)IBS-C(AUC=0.86,95%CI=0.78–0.94,P<0.0001).

A

0 20 40 60 80

0 50 100 150

B

0 20 40 60 80

0 50 100 150

C

0 20 40 60 80

0 50 100 150

D

0 20 40 60 80

0 50 100 150

Fig.8. ROCanalysisforsomatostatincelldensityexpressedasthenumberofcellspermicroscopicﬁeldin(A)IBS-total(AUC=0.93,95%CI=0.89–0.98,P<0.0001),(B)IBS-D (AUC=0.92,95%CI=0.86–0.98,P<0.0001),(C)IBS-M(AUC=0.94,95%CI=0.90–0.98,P<0.0001)and(D)IBS-C(AUC=0.93,95%CI=0.88–0.98,P<0.0001).

(7)

thehypothesisthatanxietyanddepressionaremorelikelytoresult fromthesymptomsofthedisorder[40],isnotlogical.Ultimately, thisapproachiscomplicated,timeconsuming,expensiveanddif- ﬁculttoapplyineverydayclinicalpractice.

Severalabnormalitiesin the gastrointestinalendocrine cells, includinginthestomachandthesmallandlargeintestines,have beenreportedinIBSpatients[1,29,41],andtheseabnormalities appeartoplay animportantrole inthepathophysiologyofthe disease[28,29].ThedensityofduodenalchromograninAcellshas beenproposedasabiomarkerforthediagnosisofIBS[42],andis reportedlyeffectiveatdifferentiatingIBSfromthehealthycondi- tion(AUC=97).Theﬁndingsofthepresentstudysuggesttwonew biomarkersforthediagnosisofIBS:thecelldensitiesofrectalPYY andsomatostatincells.

PYYcelldensityisagoodbiomarkerfordifferentiatingIBSfrom thehealthycondition,withAUCsof0.95and0.99whenexpressed relativetotheareaofepithelialcells andper microscopicﬁeld, respectively.WhenanalyzedaccordingtothevariousIBSsubtypes, theAUCswere 0.94, 0.99 and0.95 for thedensity of PYYcells expressedrelativetotheareasofepithelialcellsinIBS-D,IBS-M andIBS-C,respectively,and0.99,1.0and0.99,respectively,when expressedpermicroscopic ﬁeld.Thus, thedensity ofrectal PYY cellsappearstobeagoodbiomarkerforallIBSsubtypes.Can a blood-basedbiomarkerassayforcirculatingPYYandsomatostatin beused?PYYcellsoccurintheileum,colonandrectum[30,43,44].

SimilarlySomatostatincellsarelocalizedinthestomach,small-and largeintestine[15,30,43–46].Hencecirculatingsomatostatinand PYYwouldreﬂectthereleaseofthesehormonesfromthecellsin allthesesegments.TherewasnodifferencebetweenIBSpatients andcontrolsregardingfastingandpostprandialPYYplasmalev- els[47].Similarly,there wasnosigniﬁcantdifferenceinfasting somatostatinplasmalevelbetweenIBSpatientsandcontrols[48].

Thus,measuringthecirculatingPYYandsomatostatinisofnouse asamarkerforthediagnosisofIBS.

Similarly,thedensityofrectalsomatostatincellsappearstobe agoodbiomarkerfordifferentiatingbetweenIBSandthehealthy condition,withAUCsof0.86and0.93whenexpressedrelativeto theareaofepithelialcellsandpermicroscopicﬁeld,respectively.

ThisparameterisequallygoodasabiomarkerforalloftheIBS subtypes,withAUCsof0.83,0.90and0.86forIBS-D,IBS-MandIBS- C,respectively,expressedrelativetotheareaofepithelialcells,and 0.93,0.92and0.93,respectively,whenexpressedpermicroscopic ﬁeld.

ThesefindingsestablishthatPYYandsomatostatincelldensi- tiescanbeusedasbiomarkersforIBS,clearlydifferentiatingdisease fromthehealthycondition,butthequestionastowhethertheydif- ferentiateIBSfromothergastrointestinaldiseasesthatmimicIBS symptomshasyettobeanswereddefinitively.Themostcommonly missedgastrointestinaldiseasesduringsymptom-baseddiagnosis areceliacdisease,inflammatoryboweldiseasesincludingmicro- scopiccolitisandcolorectalcancer[49,50].Ithasbeenreported thatplasmalevelsofPYYareincreasedinceliacdisease[51],and thedensityoflarge-intestinalPYYcellsisincreasedinlymphocytic colitis,unchangedinulcerativecolitisandincolonandrectalcan- cers,anddecreasedinCrohn’sdisease[32,44,52,53].Ittherefore appearsthatwiththeexceptionofCrohn’sdisease,thedensityof rectalPYYcellscanpotentiallybeusedtodifferentiateIBSfrom thesegastrointestinaldiseases.Large-intestinalsomatostatincell densityisnotchangedinlymphocyticcolitisorulcerativecolitis, andisdecreasedincolonandrectalcancers[32,44,52,53].There- fore,aswithPYYcelldensity,thedensityofrectalsomatostatin cellsseemslikelytobeabletodifferentiateIBSfromsomeother gastrointestinaldiseases.

The rectum is easily accessible for biopsy sampling and is generallywelltoleratedbypatients,andimmunohistochemistry isinexpensive and usedroutinelyin all pathologylaboratories.

Furthermore,manuallycountingthePYYandsomatostatincellsin tenmicroscopicﬁeldsisrapidanddoesnotrequiresophisticated equipmentorconsiderableexperience.

Conclusion

Thisstudyhasrevealedtwonewbiomarkersforthediagnosisof IBS,namelythedensitiesoftherectalPYYandsomatostatincells.

Thesetwobiomarkersfulﬁlltherequirementsforclinically use- fulbiomarkersinthattheyhavebeenshowntopossess agood ability to differentiate IBS from the healthy condition, maybe abletodifferentiateIBSfromothergastrointestinaldiseasesthat mimicIBSsymptoms, andareinexpensive,simple, easytoper- formanddonotrequireconsiderableexperienceorsophisticated equipment.

Conﬂictsofinterest

Theauthorshavenoconﬂictsofinteresttoreport.

Authorcontributions

M.E.plannedthestudy,recruitedthepatientsandcontrolsub- jects,performedthecolonoscopyandmorphometry,analyzedthe dataanddraftedthemanuscript.J.G.H.,O.H.G.andT.H.contributed equallytotheplanningofthestudyandevaluationoftheresults, andcommenteduponthemanuscript.

Acknowledgement

ThestudywassupportedbyagrantfromHelse-Fonna(grantno.

40415).

References

[1]El-SalhyM.Irritablebowelsyndrome:diagnosisandpathogenesis.WorldJ Gastroenterol2012;18:5151–63.

[2]El-SalhyM,GundersenD,HatlebakkJG,HauskenT.Irritablebowelsyndrome:

diagnosis,pathogenesisandtreatmentoptions.NewYork:NovaSciencePub- lishers,Inc.;2012.

[3]ManningAP,ThompsonWG,HeatonKW,MorrisAF.Towardspositivediagnosis oftheirritablebowel.BrMedJ1978;2:653–4.

[4]DrossmanDA.RomeIII:thenewcriteria.ChinJDigDis2006;7:181–5.

[5]DrossmanDA.ThefunctionalgastrointestinaldisordersandtheRomeIIpro- cess.Gut1999;45(Suppl.2):II1–5.

[6]KruisW,ThiemeC,WeinzierlM,SchusslerP,HollJ,PaulusW.Adiagnosticscore fortheirritablebowelsyndrome.Itsvalueintheexclusionoforganicdisease.

Gastroenterology1984;87:1–7.

[7]BellentaniS,BaldoniP,PetrellaS,TataC,ArmocidaC,MarchegianoP,etal.A simplescorefortheidentiﬁcationofpatientsathighriskoforganicdiseases ofthecoloninthefamilydoctorconsultingroom.TheLocalIBSStudyGroup.

FamPract1990;7:307–12.

[8]Mazumdar TN, Prasad KV, Bhat PV. Formulation of a scoring chart for irritablebowelsyndrome(IBS):aprospectivestudy.IndianJGastroenterol 1988;7:101–2.

[9]LongstrethGF,ThompsonWG,CheyWD,HoughtonLA,MearinF,SpillerRC.

Functionalboweldisorders.Gastroenterology2006;130:1480–91.

[10]CamilleriM.Dothesymptom-based,Romecriteriaofirritablebowelsyndrome leadtobetterdiagnosisandtreatmentoutcomes?Theconargument.ClinGas- troenterolHepatol2009;8:129.

[11]Quigley EM. The ‘con’ case. TheRome processand functional gastroin- testinaldisorders:thebarbariansareatthegate!NeurogastroenterolMotil 2007;19:793–7.

[12]JellemaP,vanderWindtDA,SchellevisFG,vanderHorstHE.Systematicreview:

accuracyofsymptom-basedcriteriafordiagnosisofirritablebowelsyndrome inprimarycare.AlimentPharmacolTher2009;30:695–706.

[13]Wendelbo I, Mazzawi T, El-Salhy M. Increased serotonin transporter immunoreactivityintensityintheileumofpatientswithirritableboweldis- ease.MolMedRep2013;9:180–4.

[14]El-SalhyM,WendelboIH,GundersenD.ReducedchromograninAcellden- sityintheileumofpatientswithirritablebowelsyndrome.MolMedRep 2013;7:1241–4.

[15]El-SalhyM,VaaliK,DizdarV,HauskenT.Abnormalsmall-intestinalendocrine cellsinpatientswithirritablebowelsyndrome.DigDisSci2010;55:3508–13.

(8)

[16]El-SalhyM,MazzawiT,GundersenD,HauskenT.ChromograninAcellden- sityintherectumofpatientswithirritablebowelsyndrome.MolMedRep 2012;6:1223–5.

[17]El-SalhyM,Lomholt-BeckB,HauskenT.ChromograninAasapossibletoolinthe diagnosisofirritablebowelsyndrome.ScandJGastroenterol2010;45:1435–9.

[18]El-SalhyM,LilleboE,ReinemoA,SalmelidL.Ghrelininpatientswithirritable bowelsyndrome.IntJMolMed2009;23:703–7.

[19]El-SalhyM,HatlebakkJG,GundersenD,HauskenT.Endocrinecellsinthegastric oxynticmucosaofpatientswithirritablebowelsyndrome.WorldJGastrointest Endosc2014;6:176–85.

[20]El-SalhyM,GundersenD,OstgaardH,Lomholt-BeckB,HatlebakkJG,Hausken T.LowdensitiesofserotoninandpeptideYYcellsinthecolonofpatientswith irritablebowelsyndrome.DigDisSci2012;57:873–8.

[21]El-SalhyM,GundersenD,HatlebakkJG,HauskenT.Chromograninacelldensity asadiagnosticmarkerforlymphocyticcolitis.DigDisSci2012;57:3154–9.

[22]El-Salhy M, Gilja OH, Hausken T. Chromogranin A cells in the stom- ach of patients with sporadic irritable bowel syndrome. Mol Med Rep 2014;10:1753–7.

[23]SjolundK,EkmanR,WierupN.Covariationofplasmaghrelinandmotilinin irritablebowelsyndrome.Peptides2010;31:1109–12.

[24]WangSH,DongL,LuoJY,GongJ,LiL,LuXL,etal.Decreasedexpressionof serotonininthejejunumandincreasednumbersofmastcellsinthetermi- nalileuminpatientswithirritablebowelsyndrome.WorldJGastroenterol 2007;13:6041–7.

[25]Park JH, Rhee PL, Kim G, Lee JH, Kim YH, Kim JJ, et al. Enteroen- docrinecellcountscorrelatewithvisceralhypersensitivityinpatientswith diarrhoea-predominantirritablebowelsyndrome.NeurogastroenterolMotil 2006;18:539–46.

[26]CoatesMD,MahoneyCR,LindenDR,SampsonJE,ChenJ,BlaszykH,etal.Molec- ulardefectsinmucosalserotonincontentanddecreasedserotoninreuptake transporterinulcerativecolitisandirritablebowelsyndrome.Gastroenter- ology2004;126:1657–64.

[27]El-SalhyM,WendelboI,GundersenD.Serotoninandserotonintransporterin therectumofpatientswithirritableboweldisease.MolMedRep2013;8:451–5.

[28]CamilleriM.Physiologicalunderpinningsofirritablebowelsyndrome:neuro- hormonalmechanisms.JPhysiol2014;592:2967–80.

[29]El-SalhyM,GundersenD,GiljaOH,HatlebakkJG,HauskenT.Isirritablebowel syndromeanorganicdisorder.WorldJGastroenterol2014;2:384–400.

[30]El-SalhyM,GundersenD,HatlebakkJG,HauskenT.Abnormalrectalendocrine cellsinpatientswithirritablebowelsyndrome.RegulPept2014;188:60–5.

[31]el-SalhyM,StenlingR,GrimeliusL.Peptidergicinnervationandendocrinecells inthehumanliver.ScandJGastroenterol1993;28:809–15.

[32]El-SalhyM,DanielssonA,StenlingR,GrimeliusL.Colonicendocrinecellsin inﬂammatoryboweldisease.JInternMed1997;242:413–9.

[33]LinOS.Colonoscopyinirritablebowelsyndrome:whomarewereassuring.

GastrointestEndosc2005;62:900–2.

[34]Gladman LM,GorardDA.Generalpractitionerandhospitalspecialistatti- tudes to functional gastrointestinal disorders. Aliment Pharmacol Ther 2003;17:651–4.

[35]BarbaraG,StanghelliniV.BiomarkersinIBS:whenwilltheyreplacesymptoms fordiagnosisandmanagement.Gut2009;58:1571–5.

[36]CamilleriM,AndrewsCN,BharuchaAE, CarlsonPJ,Ferber I,StephensD, etal.Alterationsinexpressionofp11andSERTinmucosalbiopsyspeci- mensofpatientswithirritablebowelsyndrome.Gastroenterology2007;132:

17–25.

[37]AerssensJ,CamilleriM,TalloenW,ThielemansL,GohlmannHW,VanDen WyngaertI,etal.Alterationsinmucosalimmunityidentiﬁedinthecolonof patientswithirritablebowelsyndrome.ClinGastroenterolHepatol2008;6:

194–205.

[38]LemboAJ,NeriB,TolleyJ,BarkenD,CarrollS,PanH.Useofserumbiomark- ersinadiagnostictestforirritablebowelsyndrome.AlimentPharmacolTher 2009;29:834–42.

[39]JonesMP,CheyWD,SinghS,GongH,ShringarpureR,HoeN,etal.Abiomarker panelandpsychologicalmorbiditydifferentiatestheirritablebowelsyndrome fromhealthandprovidesnovelpathophysiologicalleads.AlimentPharmacol Ther2014;39:426–37.

[40]SpillerR,AzizQ,CreedF,EmmanuelA,HoughtonL,HunginP,etal.Guidelines ontheirritablebowelsyndrome:mechanismsandpracticalmanagement.Gut 2007;56:1770–98.

[41]El-SalhyM,HatlebakkJG,GiljaOH,HauskenT.Irritablebowelsyndrome:recent developmentsindiagnosis,pathophysiology,andtreatment.ExpertRevGas- troenterolHepatol2014;8:435–43.

[42]El-SalhyM,GiljaOH,GundersenD,HatlebakkJG,HauskenT.Duodenalchro- mograninacelldensityasabiomarkerforthediagnosisofirritablebowel syndrome.GastroenterolResPract2014;2014:462856.

[43]El-SalhyM,GiljaOH,GundersenD,HatlebakkJG,HauskenT.Endocrinecells intheileumofpatientswithirritablebowelsyndrome.WorldJGastroenterol 2014;20:2383–91.

[44]El-SalhyM,GundersenD,HatlebakkJG,HauskenT.Highdensitiesofserotonin andpeptideYYcellsinthecolonofpatientswithlymphocyticcolitis.WorldJ Gastroenterol2012;18:6070–5.

[45]El-SalhyM,GiljaOH,GundersenD,HauskenT.Endocrinecellsintheoxyntic mucosaofthestomachinpatientswithirritablebowelsyndrome.WorldJ GastrointestEndosc2014;6:176–85.

[46]El-SalhyM,GiljaOH,HatlebakkJG,HauskenT.Stomachantralendocrinecells inpatientswithirritablebowelsyndrome.IntJMolMed2014;34:967–74.

[47]VanDerVeekPP,BiemondI,MascleeAA.Proximalanddistalguthormone secretioninirritablebowelsyndrome.ScandJGastroenterol2006;41:170–7.

[48]ZhangH,YanY,ShiR,LinZ,WangM,LinL.Correlationofguthormoneswith irritablebowelsyndrome.Digestion2008;78:72–6.

[49]El-SalhyM,HalweJ,Lomholt-BeckB,GundersenD.Theprevalenceofinﬂam- matoryboweldiseases,microscopiccolitis,andcolorectalcancerinpatients withirritablebowelsyndrome.GastroenterolInsights2011;3:7–10.

[50]El-SalhyM,Lomholt-BeckB,GundersenD.Theprevalenceofceliacdiseasein patientswithirritablebowelsyndrome.MolMedRep2011;4:403–5.

[51]SjolundK,EkmanR.IncreasedplasmalevelsofpeptideYYincoeliacdisease.

ScandJGastroenterol1988;23:297–300.

[52]El-SalhyM,MahdaviJ,NorrgardO.Colonicendocrinecellsinpatientswith carcinomaofthecolon.EurJGastroenterolHepatol1998;10:517–22.

[53]El-SalhyM,NorrgårdO,FranzenL,ForsgrenS.Colonicendocrinecellsinrec- talcarcinoma,withparticularregardtopreoperativeirradiation.GICancer 1997;2:285–92.