for acute biliary pancreatitis Outcomes and timing of endoscopic retrogradecholangiopancreatography Digestive and Liver Disease

(1)

ContentslistsavailableatScienceDirect

Digestive and Liver Disease

j o ur n a l ho me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / d l d

Digestive Endoscopy

Outcomes and timing of endoscopic retrograde

cholangiopancreatography for acute biliary pancreatitis

Adrienn Halász

^a

, Dániel Pécsi

^b

, Nelli Farkas

^b,c

, Ferenc Izbéki

^a

, László Gajdán

^a

, Roland Fejes

^a

, József Hamvas

^d

, Tamás Takács

^e

, Zoltán Szepes

^e

, László Czakó

^e

, Áron Vincze

^f

, Szilárd Gódi

^f

, Andrea Szentesi

^b^,^e

, Andrea Párniczky

^g

, Dóra Illés

^e

, Balázs Kui

^e

, Péter Varjú

^b

, Katalin Márta

^b

, Márta Varga

^h

, János Novák

ⁱ

, Attila Szepes

^j

, Barnabás Bod

^k

, Miklós Ihász

^l

, Péter Hegyi

^b,m,n

, István Hritz

^o

,

Bálint Er ˝oss

^b,∗

, on behalf of the Hungarian Pancreatic Study Group

aSzentGyörgyTeachingHospitalofFejérCounty,Székesfehérvár,Hungary

bInstituteforTranslationalMedicine,MedicalSchool,UniversityofPécs,Pécs,Hungary

cInstituteofBioanalysisandInstituteforTranslationalMedicine,MedicalSchool,UniversityofPécs,Pécs,Hungary

dBajcsy-ZsilinszkyTeachingHospitalofSemmelweisUniversity,Budapest,Hungary

eFirstDepartmentofMedicine,UniversityofSzeged,Szeged,Hungary

fFirstDepartmentofMedicine,MedicalSchool,UniversityofPécs,Pécs,Hungary

gHeimPálNationalInstituteforPediatrics,Budapest,Hungary

hBMKK,Dr.RéthyPálHospital,Békéscsaba,Hungary

iBMKK,PándyKálmánHospital,Gyula,Hungary

jBács-KiskunCountyUniversityTeachingHospital,Kecskemét,Hungary

kDr.BugyiIstvánHospitalofCsongrádCounty,Szentes,Hungary

lMarkusovszkyTeachingHospital,Szombathely,Szombathely,Hungary

mFirstDepartmentofMedicine,UniversityofPécs,Pécs,Hungary

nMTA–SZTEMomentumTranslationalGastroenterologyResearchGroup,Szeged,Hungary

oFirstDepartmentofSurgery,CenterforTherapeuticEndoscopy,SemmelweisUniversity,Budapest,Hungary

a r t i c l e i n f o

Articlehistory:

Received30December2018 Accepted21March2019 Availableonline25April2019

Keywords:

Cholangiopancreatography Cholangitis

Complications Endoscopicretrograde Healthcare

Pancreatitis Qualityindicators Registries

a b s t r a c t

Background:Indicationofendoscopicretrogradecholangiopancreatography(ERCP)inacutebiliarypan- creatitis(ABP)ischallenging.

Aims:Inthisretrospectivestudy,weanalyzedreal-worlddatatounderstandtheERCPpracticeinABPin Hungariancenters.

Methods:ClinicaldataonABPpatients(2013–2015)wereextractedfromourlargemulticentricdatabase.

Outcomes,qualityindicatorsandtheroleofearlytimingofERCP(<24hfromadmission)wereanalyzed.

Results:Therewere356patientswithABP.ERCPwasperformedin267(75%).Performanceindicatorsof ERCPprovedtobesuboptimalwithabiliarycannulationrateof84%.Successfulvsunsuccessfulcannula- tionofnaïvepapillaresultedinlowerratesoflocal[22.9%vs40.9%,(P=0.012)]andsystemic[4.9%vs13.6%, (P=0.042)]complications.Successfulvsunsuccessfulclearanceresultedinlowerratesoflocalcomplica- tions[22.5%vs40.8%,(P=0.008)].Successfulcannulationanddrainagecorrelatedwithlessseverecourse ofABP[3.6%vs15.9%,(P=0.001)and4.1%vs12.2%,(P=0.033)]respectively.Atendencyofanincreased rateoflocalcomplicationswasobservedifERCPwasperformedlater[<24h:21.1%(35/166);between 24–48h:23.4%(11/47);>48h:37.2%(16/43)(P=0.088)].

Conclusion:OptimizationofERCPindicationinABPpatientsiscriticalassuboptimalERCPpracticesin ABPwithoutdeﬁnitivestonedetectionareassociatedwithpoorerclinicaloutcomes.

∗Correspondingauthorat:InstituteforTranslationalMedicine,MedicalSchool,UniversityofPécs,Szigetistr.12,Pécs,7624,Hungary.

E-mailaddresses:ahalasz@mail.fmkorhaz.hu(A.Halász),daniel.pecsi1991@gmail.com(D.Pécsi),nelli.farkas@aok.pte.hu(N.Farkas), ﬁzbeki@gmail.com(F.Izbéki),lgajdan@yahoo.com(L.Gajdán),rolldoc@vipmail.hu(R.Fejes),hamvas.jozsef@bajcsy.hu(J.Hamvas), takacs.tamas@med.u-szeged.hu(T.Takács),szepes.zoltan@med.u-szeged.hu(Z.Szepes),czako.laszlo@med.u-szeged.hu(L.Czakó),

vincze.aron@pte.hu(Á.Vincze),godi.szilard@pte.hu(S.Gódi),szentesiai@gmail.com(A.Szentesi),andrea.parniczky@gmail.com(A.Párniczky),illes.dora@med.u-szeged.hu (D.Illés),k.kubali@gmail.com(B.Kui),varjupet@gmail.com(P.Varjú),katalin.martak@gmail.com(K.Márta),drvargamarta@gmail.com(M.Varga),drnovakjanos@gmail.com (J.Novák),szepesaz@gmail.com(A.Szepes),bancikab@freemail.hu(B.Bod),ihasz.miklos@gmail.com(M.Ihász),hegyi2009@gmail.com(P.Hegyi),istvan.hritz@freemail.hu (I.Hritz),eross.balint@pte.hu(B.Er ˝oss).

https://doi.org/10.1016/j.dld.2019.03.018

(2)

Coretip

The endoscopic management of acute biliary pancreatitis is stilla controversial topic.In this prospectively collectedcohort ofpatientsinHungary,wefoundthatsuccessfulcannulationand clearanceattheﬁrstattemptareassociatedwithbetteroutcomes andsuboptimalERCPpracticesmayhaveanegativeimpactonthe outcomesofthedisease.

1. Introduction

Acutepancreatitis(AP)isoneofthemostcommondiseasesof thegastrointestinaltractrequiringacutehospitalization;itisasso- ciatedwithsigniﬁcantmorbidityandmortalityworldwidewithan increasingincidenceof5–100/100,000casesperyear[1].Despite itsimportance,researchonpancreatitisiscontinuouslydecreasing, suggestingthatmoreattentionshouldbepaidtothisdisease[2].

OneofthemainetiologicalfactorsinthepathogenesisofAP istheobstructionoftheampullaofVaterbygallstonesorsludge orbyhypertrophyof papillaandbilerefluxintothepancreatic duct,contributingto35–60%ofallAPcases.ThissubtypeofAP istermedacutebiliaryorgallstonepancreatitis(ABP)[3].Manage- mentofABPrequirestwotreatmentstrategiesinmostcases.The general,conservativemedical treatmentconsistsof appropriate fluidresuscitation(preferablylactatedRinger’ssolution)[4],pain management,andenteralnutritioninsomecases[5].Theuseofthe interventionalstrategyinABPtoachievebiliarydecompressionis stilluncertain.Incasesofconcomitantacutecholangitis,theneed forurgent(<24h)endoscopicretrogradecholangiopancreatogra- phy(ERCP)isrecommended.ThereisalsoaclearindicationofERCP incasesofobstructionwhere biliarydrainagemustbeproperly resolvedinashortperiodoftime[1,6–8].Ontheotherhand,with- outapparentsignsofcholangitisorobstruction(manifestsystemic inflammation,biliarystonesordilatationonimaging,andjaundice orabnormalliverfunctiontest),theindicationofERCPintheset- tingofABPisstilldebatedbecauseofthelackofavailableevidence [1,6].Thedataprovidedbymeta-analysesofrandomizedcontrolled trials(RCTs),wherethemostrecentonesanalyzed10and11RCTs, demonstratedasignificantdecreaseincomplications,hospitalstay, andcostinpatientswithABPmanagedwithearlyERCP(within 72h)comparedtoconservativemanagement[9,10].Anongoing RCTisorganizedbytheDutchPancreatitisStudyGrouponthisstill controversialtopic;theAPECtrialissettodeterminetheroleof earlyERCPwithbiliarysphincterotomyinABPwithoutcholangi- tis[11].Despitethestudiesnotedabove,optimalcomprehensive managementofABPisstilllackingclearevidence.

TheHungarianPancreaticStudyGroup(HPSG)wascreatedin 2011toimprove patientcareinpancreaticdiseases andwithin ashortperiod;thisorganizationhasproducedseveralpancreatic registries,trialsandestablishedguidelines[12–16].Oneofitsreg- istries,concerningAcutePancreatitis,hasrecordeddataonAPcases fromallparticipatingcentersthroughoutHungaryandnowinmore thanadozencountries[17].

ThisstudyaimstodeterminetheroleofERCPandsphinctero- tomyandstoneclearanceinABPandtoprovideanoverviewofthe generaluseofERCPinHungaryinthisdisease.

2. Materialsandmethods 2.1. Inclusioncriteria

AllpatientswithAPwereenrolled,andtheirdatawereprospec- tivelycollectedintheHPSGAPRegistry,whichhasbeenapproved bytheScientiﬁcand ResearchEthics Committeeof theMedical ResearchCouncil (TUKEB-22254-1/2012/EKU).Allpatientswere

informedaboutthedatacollectionandsignedtheinformedconsent forms.AnAPdiagnosiswasmadeaccordingtotherecommenda- tionsintheIAP/APAguidelines,withatleasttwoofthefollowing threecriteriamet:abdominalpain,pancreaticenzymeexceeding morethanthreetimestheuppernormallevel,andfeaturesofpan- creatitisonimaging.

Inthiscohortstudy,weselectedpatientswhoﬁtthecriteria previouslylaiddownbytheDutchPancreaticStudyGroup,which wereusedtodeterminebiliaryorigin:(a)gallstonesand/orsludge diagnosedontransabdominalultrasoundorcomputedtomography (CT)or(b)dilatedCBDonultrasoundorCT(diameter:>8mmfor age≤75yearsanddiameter:>10mmforage>75years)or(c)two ofthefollowingthreelaboratoryabnormalities:(1)serumbilirubin level>1.3mg/dL[>40␮mol/L];(2)alanineaminotransferase(ALT) level>100U/LwithanALATlevelgreaterthantheaspartateamino- transferaselevel;and(3)alkalinephosphataselevel>195U/Lwith agamma-glutamyltransferaselevel>45U/L.OthercausesofAP, suchasalcohol,hypertriglyceridemia,diet,drug-induced,trauma, viralinfection,post-ERCP,andidiopathicAPhadtobeabsent(Sup- plementaryTable1)[18].

691patientswithAPwereenrolledintheAPregistrybetween January2013andAugust2015from14centers.Themanuscript waspreparedinaccordancewiththeSTROBEstatement[19].

2.2. Exclusioncriteria

Patientsundertheageof18yearsandthosewithnon-biliary pancreatitiswereexcludedfromtheanalysis.

2.3. Dataextraction

Dataondemographics(sexandage),etiology,severity,andmor- talityofAPwereextractedforallsubjectswithAP,andadescriptive statisticalanalysiswasperformed.TheseverityofAPwasclassified accordingtotherevisedAtlantaclassificationasmild,moderately severe,andsevere[20].Mainoutcomesweretheseverityofpancre- atitis,local(peripancreaticfluid,pseudocyst,necrosisofpancreas onimaging,diabetesmellitus,andabdominalcompartmentsyn- drome)andsystemic(transientorpersistentorganfailurebased onthemodifiedMarshallscoringsystemfororgandysfunction) complications,mortality,andlengthofhospitalstay.

Detailed demographics, including body mass index, co- morbidities,and dataonoutcomesfor ERCP, werecollectedon patientswithABP,suchasindicationofERCP,successfulcannula- tionrate,managementofCBDstonesbysphincterotomyandduct clearance,biliaryandpancreaticstenting,anatomyofthepapilla (naïve/notnaïve),andcomplicationrates(bleedingandperfora- tion).TimingofERCPwascalculatedfromadmission.Theoutcomes forABP(severityofpancreatitis,localandsystemiccomplications, mortality,andlengthofhospitalstay)wereanalyzedinrelation tothetimingofERCP.Nofollow-upwascarriedoutafterhospital discharge.

2.4. Statisticalanalysis

Continuousmeasuresaresummarizedandpresentedasmeans andstandarddeviations(SD)orasmedianandinterquartileranges (IQR).Categoricaldataarepresentedasobservedandaspercent- ages.Todeterminedifferencesbetweencontinuousparameters, depending on the distribution of the data, we used the inde- pendent Student’s t-test or the Mann–Whitney U test for two groups and one-way ANOVAwiththe Bonferroni post-hoc test orKruskal–Wallistest incomparingmorethantwo groups.We usedtheChi-squaretestorFisher’sexacttesttoanalyzetherela-

(3)

Table1

Generalcharacteristicsoftheacutebiliarypancreatitiscohort.

All[n(%)] Women[n(%)] Men[n(%)] P-value

Gender 356 204(57.3%) 152(42.7%) –

Age 61.65±17.32 61.3±18.1 62.1±16.1 NS

Priorcholecystectomy 35(9.4%) 28(13.9%) 7(4.6%) 0.004

Previouslydocumentedpancreatitis 42(11.8%) 26(12.7%) 16(10.5%) NS

Diabetesmellitus 59(16.6%) 26(12.7%) 33(21.7%) 0.023

Bodymassindex(availablefor160womenand123men) 28.31±6.1 28.24±6.03 28.40±6.22 NS

Atleast2co-morbidities 155(43.5%) 85/204(41.7%) 70/152(46.1%) NS

tionsbetweenthefactorsunderexamination.Allanalyseswere performedwithSPSS24statisticalsoftware(IBMCorporation).

3. Results

3.1. GeneralcharacteristicsoftheAPcohort

Biliary etiology wasfound in 356 (51.5%) patients, and 335 (48.5%)patientshadotheretiologicalfactors(alcohol,hypertriglyceridemia,diet,drug-induced,trauma,viralinfection,post-ERCP, and idiopathic AP). Among the subjects with ABP, there were more women, and they were older than patients with a dif- ferent etiology [204/356(57.3%) vs 106/335(31.6%) (P<0.001)]

andmeanage[61.5±17.32vs51.47±15.73years(P<0.001)].The courseof pancreatitiswith biliaryetiology wasmilder in con- trasttonon-ABPdisease[mildABP:248/356(69.7%)vsnon-ABP:

183/335(54.6%);moderatelysevereABP:86/356(24.2%)vsnon- ABP: 121/335 (36.1%); severe ABP: 22/356 (6.2%) vs non-ABP:

31/335 (9.2%) (P<0.001)].Therewasno differencein mortality betweenthetwogroups[ABP:8/336(2.4%)vsnon-ABP:13/322 (4.0%)(P=0.242)].

3.2. CharacteristicsoftheABPcohort

Apancreatitisdiagnosiswasbasedonupperabdominalpainand elevatedpancreaticenzymesin327/356(91.8%)ofthepatients.

ABPoccursmorecommonlyinwomen[204/356(57.3%)female vs152/356(42.7%)male].Inalmost10%ofthecases,ABPdevel- opedafteracholecystectomy[35/356(9.4%)]andmorefrequently inwomen[28/204(13.9%)vs7/152(4.6%)(P=0.004)].

Diabetesmellitusasaco-morbiditywasfoundin16.6%(59/356) ofthepatients,signiﬁcantlymoreofteninmen[26/204(12.7%)vs 33/152(21.7%)(P=0.023)].

Age, a previously documented episode of pancreatitis, body massindex(BMI),andco-morbiditywerenotdifferentbetween thetwosexes.AconsiderablenumberofABPpatientshadmore thantwoco-morbidities[43.5%(155/356)](Table1).

3.3. IndicationsforERCP

Outofthe356patients,267underwentERCP(75.0%)forsus- pected cholangitis or cholestasis without cholangitis based on raisedinﬂammatorymarkerswithdilatedbiliaryductsandraised liver function tests. 89 patients in total did not undergo ERCP althoughitwouldhave beenindicated in50casesofsuspected cholangitis(56.2%)ERCPwasnotperformedinthesecasesdueto animprovingclinicalpicture,lackofconsentfromthepatient,or rapiddeteriorationofmulti-organfailure.

Endoscopicultrasonography(EUS)wasperformedinonlyﬁve patients,withbileductstonesbeingidentiﬁedintwocases.MRCP wascarried out in one patient, in which clear bile ductswere reported.

3.4. QualityindicatorsandﬁndingsofERCP

ThekeyperformanceindicatorsforERCPmetthecriteriaset outintheAmericanSocietyofGastrointestinalEndoscopy(ASGE) guidelines[21].Successfulbiliarycannulationwasachievedin233 subjectswithnaïvepapilla(90.7%),butthesuccessfulcannulation ratewas84.0%(216procedures)atﬁrstattempt.In80subjects, extractionsofstonessmallerthan1cmweresuccessfulin93.7%

ofthecases.Stentimplantationbelowthebifurcationwassuccess- fullycarriedoutinallcasesaftersuccessfuldeepbiliarycannulation (33/33).Perforationoccurredin1/267(0.4%)ofthecases.Clinically signiﬁcantbleedingrequiringbloodtransfusiondevelopedin3/267 (1.2%)ofthepatients.

Commonbileduct(CBD)stones,sludge,and/ordilationofthe bileductswerereportedin97(36.3%),91(34.1%),and124(46.4%) cases,respectively.Spontaneouspassageofabileductstonewas suspectedin19.5%(52/267)ofthepatientsduringERCP.In30cases (11.2%),nobiliarypathologywasfoundbyERCP.Endoscopicultra- soundwasonlycarriedoutinﬁvecasesbecauseoflimitedaccess atthetimeofdatacollection.

315ERCPswereperformedin267patientsuntilcompletion orabandoningtheinterventionortreatment.43patientshadtwo ERCPs,andﬁvehadthreeprocedures.

3.5. OutcomesforABPinrelationtosuccessratesofERCP

Dataoncannulationsuccessrateandclearanceofthebileducts wereavailableinallcases.Thesuccessrateforbileductcannulation inallpatientswas83.5%(223/267)duringtheﬁrstERCP,andany furtherendoscopicattemptsresultedinahigherrateofsuccess [90.6%(242/267)].Successfulcannulationwasachievedin84.0%

(216/257)ofpatientswithnaïvepapilla,andclearanceofthebile ductwassuccessfulin71.5%(191/267)attheﬁrstERCPattempt.

Endoscopicbiliarysphincterotomywasdonein86.5%(231/267)of theERCPs,whereaspancreaticsphincterotomywasonlyperformed in1.12%(3/267)ofthecases.Biliarystentswereplacedin12.36%

(33/267)andpancreaticstentsin16.85%(45/267)ofthecases.Suc- cessfulcannulationwasassociatedwithsigniﬁcantlylowerratesof localandsystemiccomplications.Successfulclearancewaslinked tolowerratesoflocalcomplications.Successfulcannulationand clearance bothcorrelated witha less severecourseof ABPand shorterhospitalization(Tables2and3).

Completefailure ofclearance anddecompression ofthebile ductswererelatedtohigherfrequencyoflocalcomplicationsand amoreseverecourseofABPandlongerhospitalstay(Table4).

3.6. OutcomesforABPinrelationtothetimingofERCP

ERCPwasperformedin75%(267/356)ofthecases,themajority ofthemduringtheﬁrst24hafteradmission.

DataonthetimingofERCPwereavailablein256(95.9%)cases.

ERCPwasperformedon64.8%(166/256)ofthepatientswithin24h afteradmission,in18.4%(47/256)ofthembetween24and48h afteradmission,andin16.8%(43/256)caseslaterthan48hafter admission.Atendencyofanincreasedrateoflocalcomplications

(4)

Table2

Successfulcannulationversusfailuretocannulatethecommonbileductwiththeﬁrstendoscopicretrogradecholangiopancreatographyintheacutebiliarypancreatitis cohort.

Successfulbileductcannulationwith1stERCP[n(%)] Failuretocannulatethebileductwith1stERCP[n(%)] P-value

Total(n=267) 223(83.5%) 44(16.5%) –

Ratesofseveredisease 8(3.6%) 7(15.9%) 0.001

Localcomplications 51(22.9%) 18(40.9%) 0.012

Systemiccomplications 11(4.9%) 6(13.6%) 0.042

Mortality 4(1.8%) 1(2.3%) NS

Hospitalstay,median(IQR) 9(6–13) 14(8–21.5) 0.00021

Table3

Successfulclearanceversusfailuretoachieveclearanceofthecommonbileductforallendoscopicretrogradecholangiopancreatographiesintheacutebiliarypancreatitis cohort.

Successfulbileductclearance[n(%)] Failureofclearancethebileduct[n(%)] P-value

Total(n=267) 218(81.6%) 49(18.3%) –

Systemiccomplications 13(6.0%) 4(8.2%) NS

Mortality 4(1.8%) 1(2.0%) NS

Hospitalstay,median(IQR) 9(6–13) 11(7–21) 0.021

Table4

Successfulclearanceversusfailureofclearanceanddecompressionofcommonbileductforallendoscopicretrogradecholangiopancreatographiesintheacutebiliary pancreatitiscohort.

Successfulclearanceofbileduct[n(%)] Unsuccessfulclearanceanddecompressionofbileduct[n(%)] P-value

Total 218/267(81.6%) 32/250(12.8%) –

Systemiccomplications 13(6.0%) 4(12.5%) NS

Mortality 4(1.8%) 1(3.1%) NS

Hospitalstaymedian(IQR) 9(6–13) 16(8.5–24.5) 0.001

Table5

Outcomesforacutebiliarypancreatitisinrelationtothetimingofendoscopicretrogradecholangiopancreatographyinallpatientswithacutebiliarypancreatitis.

<24h[n(%)] 24–48h[n(%)] >48h[n(%)] P-value

Total(n=256) 166(64.8%) 47(18.36%) 43(16.80%) –

Ratesofseveredisease 5(3%) 3(6.4%) 3(7.0%) NS

Localcomplications 35(21.1%) 11(23.4%) 16(37.2%) 0.088

Systemiccomplications 7(4.2%) 4(8.5%) 4(9.3%) N/A^*

Mortality 2(1.2%) 1(2.1%) 0(0%) N/A^*

Hospitalstay,median(IQR) 8(6–12) 10(5.5–15) 13(9.5–21) <0.001

* Statisticalanalysiswasnotcarriedoutincasesofsystemiccomplications,mortality,andratesofseverediseaseduetolownumbersofsubjects.

wasobservedifERCP wasperformedlater[ERCPin24h:21.1%

(35/166);between24and48h:23.4%(11/47);after48h:37.2%

(16/43)(P=0.088)].

Wenotethatonly1.1%(7/267)oftheendoscopicinterventions describedsignsofpurulentcholangitis.

The length of hospitalization was signiﬁcantly longer in all patientsifERCPwasdelayed(Table5).

3.7. Otherrelevantﬁndings

Therewasnostatisticallysigniﬁcantdifferenceintheoutcomes forABPbetweenthepatientstreatedwithorwithoutERCP.(Sup- plementaryTable2).

Useofantibioticswasacommonpractice.87.6%(312/356)of allthepatientsreceivedantibiotics,forwhichtheindicationwas suspectedcholangitisin85.3%(266/312).Theywereadministered totreatinfectionsoutsidethebiliarytree,suchaspneumoniaand urinarytractinfections, in11.9%(37/312)of thecases.Theﬁrst choiceamongantibioticsforcholangitiswasacombination ofa cephalosporinandmetronidazole.

4. Discussion

Prospectivelycollected,real-worlddatawereanalyzedinthis multicenterstudy,andconsiderablecoverageofHungarianacute biliarypancreatitiscasesispresented.Datashown abovedepict currentmanagementstrategiesusedinHungary.

Although a clear-cut diagnosisof definite cholangitis would have beendesirable in our analysis,currently there is novali- dated definitionof cholangitisinthe settingof ABP.Simple AP canresultinatransientandself-resolvingbiliaryobstructionwith derangedliverfunctiontestsanddilatedbiliarytree,whichcan mimiccholangitiswiththeraisedinflammatorymarkersdrivenby pancreatitis.Thustheuseofthedefinitionofdefinitecholangitisas termedbytheTokyocriteriahadtobeavoided[18].

Asdescribedinotherstudies,wefoundthatpatientswithABP areolder,andtherearemorewomenamongthemcomparedtoAP ofotheretiologies[3].IntheHungariancohort,ABPtendedtohave alessseverenaturalcourse,butthemortalitywasthesameasin otheretiologies,asreportedinalargestudy[2].

Previouscholecystectomywasrelativelycommonandmoreso inwomen.Thiscouldbeexplainedbythefactthatbiliarystone diseaseismorecommoninfemalesandthatstonediseaseofthe

(5)

gallbladderincreasestheriskofABP,mostlikelyevenafterapre- viouscholecystectomy.Apreviously documentedepisode ofAP occurredinmorethan10%ofthepatients.Althoughdataonthe etiologyofthepreviousattackwasnotavailable,webelievethat themajorityofthecaseswerelikelydrivenbygallstonedisease, similartodatareportedbyGodietal.[12].Diabetesisaknown riskfactorforAP,andsigniﬁcantlymoremenhaddiabetesinthe Hungariancohort,whichwasreportedinAPwithalletiologies[15].

ERCPwasperformedin75%ofthepatientspresentingwithABP.

Toourbestknowledge,therearenopreviouscohortstudieswhere therateofERCPwaspublishedandanalyzed.Wefoundthatour ERCPpracticeinABPisinlinewiththecurrentguidelines;however, wemusthighlightthatverylimitedaccesstourgentendoscopic ultrasound(EUS)andmagneticresonancecholangiopancreatogra- phy(MRCP)resultedinanumberofavoidableERCPs.Atthesame time,asmallproportionofthepatientswithsuspectedcholangitis werenotamenabletoERCP.Insummary,webelievethatERCPfor ABPwillbereducedasaccesstoEUSandMRCPimproves.Inmost cases,theindicationwassuspectedcholangitis,andthesearethe patientswhocouldhavebeneﬁtedmostfromadditionaldiagnostic imaging[16,19,22].

Ourresultsclearlydemonstratedthatthelackofaccesstoaddi- tionaldiagnostictools(EUSorMRCP)resultedinahighnumberof unnecessaryandavoidableERCPs,andthisclinicalpracticeneeds tobeimproved.

Someof thekeyperformance indicatorsdescribingtheERCP practicesinthislargecohortacrossmanycentersdescribedsub- optimalERCPpractices.Mostimportantlyasuccessrateof84%at ﬁrstattempt(216patients),whichisbelowthequalitybenchmark of>90%recommendedbyASGE.Thismaywellbedrivenbythe factthatsomeoftheERCPswereperformedinlowvolumecenters.

Italsoremindsusthat,ifindicated,high-qualityERCPwithmaxi- malpancreasprotectionandhighcompetenceofalternativebiliary accesstechniquesshouldbemandatory.

Oneofourmainfindingsisthatfailedcannulationandbileduct clearanceareassociatedwithahigherincidenceoflocalcompli- cationsandseverityofABP.Thisresultcanbeinterpretedintwo ways.Firstly,successfulclearanceanddecompressionofthebile ductscanresultinaquickerresolutionofpancreatitisandlesspro- gressionleadingtocomplications.Secondly,itmaybeexplained bythedifficultaccesstothebileductsinalreadycomplicatedAP, drivenbydifficultintubationoftheduodenum,poorvisualization ofthepapilla,limitedmaneuverabilityoftheduodenoscope,and challengingcannulationoftheedematouspapilla.ERCPsarethere- foredonefortheindicationofacutebiliarypancreatitisclassified asgrade3difficultyonthemodifiedSchutzgrade,onascaleof1–4, where4isthemostdifficult[23].

Inthissituation,highsuccessratescanonlybeexpectedofcom- petent,highlyskilledendoscopistswithsubstantialcasenumbers.

Thisishowweexplaintheslightlysuboptimalqualityindicators (ASGEguideline)ofERCPinthiscohort[19].Cannulationofnaïve papillawassuccessfulatﬁrstattemptin84.1%ofallERCPs(desired:

90%),perforationoccurredin0.4%(desired:≤0.2%),andbleeding requiringtransfusionresultedin1.2%(desired:≤1%).Wenotethat thisanalysiscontaineddatafrom267patients,hencethetwolat- termeasures.Qualityindicatorsofstoneextractionandstentingof obstructionsbelowthelevelofbifurcationmetthecriteriaforthe guidelines.

Wedidnotfindasignificantdecreaseintherateoflocalcompli- cationsandhospitalstayinthecohortwhenERCPwasperformed withintwodays.EvidencesuggeststhatearlyERCPinABPwith cholangitisis indicated[1,6–8],but ourfindingscouldnotrein- forcethesepreviousdata.Inpatientswithaclear-cutdiagnosisof acutecholangitis,ERCPshouldbeconsideredassoonaspossibleto provideabetteroutcome[24].

Just like the need for ERCP, the high rate of antibiotic use reportedinourstudy(87.6%) couldpossiblybereducedbybet- teraccesstoEUSandMRCPinthecaseofsuspectedcholangitis.

However,cholangitisisoneofthemostfearedsourcesofabdom- inalinfectionandcanleadtosepsis,multipleorganfailure,and death.Therefore,anystrategytodelayorwithholdantibioticsinthe contextofsuspectedcholangitisshouldbecarefullyassessed.Edu- cationalactivitiesshouldbeorganizedandmaterialsdisseminated toensurestrictadherencetointernationalguidelines[1,6].

4.1. Strengthsandweaknessesofthestudy

Thiscohortrepresentsageneral,diverse,multicenter(notonly tertiary centers participated), acute biliary pancreatitissample.

Forthisreason,broader,moregeneralizableconclusionscouldbe drawn.Limitationsofthisstudyaretherelativelylowcasenumbers insubgroupsandtheretrospectivedesignwithpost-hocquestion raising,whichissusceptibletobiases,thuslimitingtheconclusions considerably.Lastly, alargenumber ofparticipatingERCPunits withoutastructuredapproachtothetimingofERCPprocedures inABPlimitsthestatisticalconclusionsonassociationswiththe outcomesofpancreatitis.

5. Conclusion

Theindicationand beneﬁtof ERCPin patientswithABPbut withoutaclear-cutdiagnosis,cholangitisremainsa contentious issue.Werecommendthatthenon-invasivediagnosticapproach shouldbemaximizedtoselectthemostsuitablesubgroupofthese patients. AsERCP isdifﬁcultin ABP,quality indicatorsmust be closelymonitored,andproceduresshouldbeperformedbyexperts inhigh-volumecentersassuboptimalERCPpracticesarelikelytobe associatedwithpooreroutcomesoftheacutebiliarypancreatitis.

Funding

Thisstudy wassupportedbyProject Grants(KH125678 and K116634 to PH, K120335 to TT), the Economic Development andInnovationOperativeProgrammeGrant(GINOP2.3.2-15-372 2016-00048toPH)and Human ResourcesDevelopment Opera- tionalProgrammeGrant(EFOP-3.6.2-16-2017-00006toPH)from the National Research, Development and Innovation Ofﬁce, by a Momentum Grant from the Hungarian Academy of Sciences (LP2014-10/2014toPH),bytheJánosBolyaiResearchScholarship of the Hungarian Academy of Sciences (to AP) and the ÚNKP- 18-4newnationalexcellenceprogramoftheMinistryofHuman Capacities(toAP).Fundershadnoroleinstudydesign,datacollec- tion,analysis,interpretationoftheﬁndings,andpreparationofthe manuscript.

Institutionalreviewboardstatement

This study was approved by the Hungarian Scientiﬁc and Research Committee of Medical Research Council (MRC). All patientswithacutepancreatitiswereenrolled,andtheirdatawere prospectivelycollectedintheHungarianPancreaticStudyGroup (HPSG)AP Registry,which hasbeen approvedby theScientiﬁc andResearchEthicsCommitteeoftheMedical ResearchCouncil (TUKEB-22254-1/2012/EKU).Allpatientswereinformedaboutthe datacollectionandsignedtheinformedconsentforms.

Conﬂictofinterest

AdriennHalászMD,DánielPécsiMD,NelliFarkasMScPhD,Ferenc IzbékiMDPhD,LászlóGajdánMD,RolandFejesMD,JózsefHam- vasMD,TamásTakácsMDPhDDSc,ZoltánSzepesMDPhD,László CzakóMDPhDDSc,ÁronVinczeMDPhDDSc,SzilárdGódiMD,

(6)

AndreaSzentesiMScPhD,AndreaPárniczkyMDPhD,DóraIllés MD,BalázsKuiMDPhD,PéterVarjúMD,KatalinMártaMD,Márta VargaMD,JánosNovákMD,AttilaSzepesMDPhD,BarnabásBod MD,MiklósIhászMD,PéterHegyiMDPhDDSc,IstvánHritzMDPhD andBálintEr ˝ossMDhaveparticipatedin(a)conceptionanddesign, oranalysisandinterpretationofthedata;(b)draftingthearticle orrevisingitcriticallyforimportantintellectualcontent;and(c) approvalofthefinalversion.Themanuscripthasnotbeensubmit- tedto,norisunderreviewat,anotherjournalorotherpublishing venue.Theauthorshavenoaffiliationwithanyorganizationwitha directorindirectfinancialinterestinthesubjectmatterdiscussed inthemanuscript.

AppendixA. Supplementarydata

Supplementarymaterialrelated tothis article canbefound, intheonlineversion,atdoi:https://doi.org/10.1016/j.dld.2019.03.

018.

References

[1]CrockettSD,WaniS,GardnerTB,Falck-YtterY,BarkunAN.AmericanGas- troenterologicalAssociationInstituteguidelineoninitialmanagementofacute pancreatitis.Gastroenterology2018;154:1096–101.

[2]SzentesiA,TothE,BalintE,FanczalJ,MadácsyT,LaczkóD,etal.Analysisof researchactivityingastroenterology:pancreatitisisinrealdanger.PLoSOne 2016;11:e0165244.

[3]vanGeenenEJ,vanderPeetDL,BhagirathP,MulderCJ,BrunoMJ.Etiol- ogyanddiagnosisofacutebiliarypancreatitis.NatRevGastroenterolHepatol 2010;7:495–502.

[4]de-MadariaE,Herrera-MaranteI,Gonzalez-CamachoV,BonjochL,Quesada- VázquezN,Almenta-SaavedraI,etal.FluidresuscitationwithlactatedRinger’s solutionvsnormalsalineinacutepancreatitis:Atriple-blind,randomized, controlledtrial.UnitedEuropeanGastroenterolJ2018;6:63–72.

[5]VaughnVM,ShusterD,RogersMAM,MannJ,ConteML,SaintS,etal.Early versusdelayedfeedinginpatientswithacutepancreatitis:asystematicreview.

AnnInternMed2017;166:883–92.

[6]IAP/APAevidence-basedguidelinesforthemanagementofacutepancreatitis.

Pancreatology2013;13:e1–15.

[7]YokoeM,TakadaT,MayumiT,YoshidaM,IsajiS,WadaK,etal.Japaneseguide- linesforthemanagementofacutepancreatitis:JapaneseGuidelines2015.J HepatobiliaryPancreatSci2015;22:405–32.

[8]MiuraF,OkamotoK,TakadaT,StrasbergSM,AsbunHJ,PittHA,etal.Tokyo Guidelines2018:initialmanagementofacutebiliaryinfectionandﬂowchart foracutecholangitis.JHepatobiliaryPancreatSci2018;25:31–40.

[9]CoutinhoLMA,BernardoWM,RochaRS,MarinhoFR,DelgadoA,MouraETH, etal.Earlyendoscopicretrogradecholangiopancreatographyversusconserva-

tivetreatmentinpatientswithacutebiliarypancreatitis:systematicreview andmeta-analysisofrandomizedcontrolledtrials.Pancreas2018;47:444–53.

[10]Burstow MJ, Yunus RM, Hossain MB, Khan S, Memon B, Memon MA.

Meta-analysis of early endoscopic retrograde cholangiopancreatography (ERCP) +/- endoscopic sphincterotomy (ES) versusconservative manage- mentforgallstonepancreatitis(GSP).SurgLaparoscEndoscPercutanTech 2015;25:185–203.

[11]SchepersNJ,BakkerOJ,BesselinkMG,BollenTL,DijkgraafMG,vanEijckCH, etal.Earlybiliarydecompressionversusconservativetreatmentinacutebiliary pancreatitis(APECtrial):studyprotocolforarandomizedcontrolledtrial.Trials 2016;17:5.

[12]GodiS,ErossB,GyomberZ,SzentesiA,FarkasN,ParniczkyA,etal.Centralized careforacutepancreatitissigniﬁcantlyimprovesoutcomes.JGastrointestin LiverDis2018;27:151–7.

[13]SzucsA,MarjaiT,SzentesiA,FarkasN,ParniczkyA,NagyG,etal.Chronicpan- creatitis:multicentreprospectivedatacollectionandanalysisbytheHungarian PancreaticStudyGroup.PLoSOne2017;12:e0171420.

[14]MartaK,SzaboAN,PecsiD,VarjuP,BajorJ,GodiS,etal.Highversuslow energyadministrationintheearlyphaseofacutepancreatitis(GOULASHtrial):

protocolofamulticentrerandomiseddouble-blindclinicaltrial.BMJOpen 2017;7:e015874.

[15]ParniczkyA,Abu-El-HaijaM,HusainS,LoweM,OraczG,Sahin-TóthM,etal.

EPC/HPSGevidence-basedguidelinesforthemanagementofpediatricpancre- atitis.Pancreatology2018;18:146–60.

[16]LakatosG,BalazsA,KuiB,GodiS,SzucsA,SzentesiA,etal.Pancreaticcan- cer:multicenterprospectivedatacollectionandanalysisbytheHungarian PancreaticStudyGroup.JGastrointestinLiverDis2016;25:219–25.

[17]ParniczkyA,KuiB,SzentesiA,BalazsA,SzucsA,MosztbacherD,etal.Prospec- tive,multicentre,nationwideclinicaldatafrom600casesofacutepancreatitis.

PLoSOne2016;11:e0165309.

[18]vanSantvoortHC,BesselinkMG,deVriesAC,BoermeesterMA,FischerK,Bollen TL,etal.Earlyendoscopicretrogradecholangiopancreatographyinpredicted severeacutebiliarypancreatitis:aprospectivemulticenterstudy.AnnSurg 2009;250:68–75.

[19]vonElmE,AltmanDG,EggerM,PocockSJ,GotzschePC,Vandenbroucke JP. StrengtheningtheReportingofObservationalStudiesinEpidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ 2007;335:806–8.

[20]BanksPA,BollenTL,DervenisC,GooszenHG,JohnsonCD,SarrMG,etal.Clas- sificationofacutepancreatitis–2012:revisionoftheAtlantaclassificationand definitionsbyinternationalconsensus.Gut2013;62:102–11.

[21]BaronTH,PetersenBT,MergenerK,ChakA,CohenJ,DealSE,etal.Quality indicatorsforendoscopicretrogradecholangiopancreatography.AmJGas- troenterol2006;101:892–7.

[22]LiuCL,FanST,LoCM,TsoWK,WongY,PoonRT,etal.Comparisonofearlyendo- scopicultrasonographyandendoscopicretrogradecholangiopancreatography inthemanagementofacutebiliarypancreatitis:aprospectiverandomized study.ClinGastroenterolHepatol2005;3:1238–44.

[23]CottonPB,EisenG,RomagnuoloJ,VargoJ,BaronT,TarnaskyP,etal.Grading thecomplexityofendoscopicprocedures:resultsofanASGEworkingparty.

GastrointestEndosc2011;73:868–74.

[24]ParikhMP,WadhwaV,ThotaPN,LopezR,SanakaMR.Outcomesassociated withtimingofERCPinacutecholangitissecondarytocholedocholithiasis.J ClinGastroenterol2018;52(10):e97–102.