Transluminal endoscopic step-up approach versus minimally invasive surgical step-up approach in ...

13 Pages·2013·1.96 MB·English

vanBrunschotetal.BMCGastroenterology2013,13:161
http://www.biomedcentral.com/1471-230X/13/161
STUDY PROTOCOL Open Access
Transluminal endoscopic step-up approach versus
minimally invasive surgical step-up approach in
patients with infected necrotising pancreatitis
(TENSION trial): design and rationale of a
randomised controlled multicenter trial
[ISRCTN09186711]
Sandra van Brunschot1,2*, Janneke van Grinsven1,2, Rogier P Voermans1, Olaf J Bakker3, Marc GH Besselink4,
Marja A Boermeester4, Thomas L Bollen5, Koop Bosscha6, Stefan A Bouwense2,7, Marco J Bruno8,
Vincent C Cappendijk9, Esther C Consten10, Cornelis H Dejong11, Marcel GW Dijkgraaf12, Casper H van Eijck13,
G Willemien Erkelens14, Harry van Goor7, Mohammed Hadithi15, Jan-Willem Haveman16, Sijbrand H Hofker16,
Jeroen JM Jansen17, Johan S Laméris18, Krijn P van Lienden18, Eric R Manusama19, Maarten A Meijssen20,
Chris J Mulder21, Vincent B Nieuwenhuis22, Jan-Werner Poley8, Rogier J de Ridder23, Camiel Rosman24,
Alexander F Schaapherder25, Joris J Scheepers26, Erik J Schoon27, Tom Seerden28, BW Marcel Spanier29,
Jan Willem A Straathof30, Robin Timmer31, Niels G Venneman32, Frank P Vleggaar33, Ben J Witteman34,
Hein G Gooszen2, Hjalmar C van Santvoort3, and Paul Fockens1 for the Dutch Pancreatitis Study Group
Abstract
Background: Infected necrotising pancreatitis is a potentially lethal disease that nearly always requires intervention.
Traditionally, primary opennecrosectomy has been the treatment ofchoice. Inrecentyears, thesurgical step-up
approach, consisting ofpercutaneous catheter drainage followed, if necessary, by (minimally invasive) surgical
necrosectomy has become the standard of care. A promising minimally invasive alternativeis theendoscopic
transluminal step-upapproach. This approach consists ofendoscopic transluminal drainage followed, if necessary,
byendoscopic transluminal necrosectomy.We hypothesisethattheless invasive endoscopic step-up approach is
superior to thesurgical step-up approach in terms of clinical and economic outcomes.
(Continuedonnextpage)
*Correspondence:[email protected]
1DepartmentofGastroenterologyandHepatology,UniversityofAmsterdam,
Amsterdam,TheNetherlands
2DepartmentofOR/EvidenceBasedSurgery,RadboudUniversityNijmegen
MedicalCentre,Nijmegen,TheNetherlands
Fulllistofauthorinformationisavailableattheendofthearticle
©2013vanBrunschotetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsofthe
CreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,
distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.TheCreativeCommonsPublic
DomainDedicationwaiver(http://creativecommons.org/publicdomain/zero/1.0/)appliestothedatamadeavailableinthis
article,unlessotherwisestated.
vanBrunschotetal.BMCGastroenterology2013,13:161 Page2of13
http://www.biomedcentral.com/1471-230X/13/161
(Continuedfrompreviouspage)
Methods/Design:The TENSION trial is a randomised controlled, parallel-group superiority multicenter trial. Patients
with(suspected) infected necrotising pancreatitis with anindication for intervention and in whomboth treatment
modalities are deemed possible, will be randomised to either an endoscopic transluminal or a surgical step-up
approach. During a 4 year study period, 98 patients will be enrolled from 24 hospitals of the Dutch Pancreatitis
Study Group. The primary endpoint is a composite of death and majorcomplicationswithin 6 months following
randomisation. Secondary endpoints include complicationssuch as pancreaticocutaneous fistula, exocrine or
endocrine pancreatic insufficiency, need for additional radiological, endoscopic or surgical intervention, theneed for
necrosectomy after drainage, thenumber of (re-)interventions,quality oflife,and total direct and indirect costs.
Discussion: The TENSION trial will answer the question whether an endoscopic step-up approach reduces the
combined primary endpoint of death and major complications,as well as hospital stay and related costs compared
witha surgical step-up approach in patients withinfectednecrotising pancreatitis.
Keywords: Acute pancreatitis, Necrotising, Treatment,Drainage, Trial, Endoscopy, Minimally invasive, Surgery,
Necrosectomy, Pancreas
Background smallrandomisedpilottrialhasshownthatendoscopic
Acute pancreatitis is a common and potentially lethal necrosectomy is feasible and reduces the inflammatory
disease. About 20% of patients develop necrosis of the response, and possibly complications such as new onset
pancreatic parenchyma and/or extrapancreatic fat tis- organ failure compared with surgical necrosectomy in
sue [1,2]. Necrotising pancreatitis is associated with these often already critically ill patients [8]. Although ini-
pancreatic and/or peripancreatic collections with fluid tial results seem promising, a randomised controlled trial
andnecrosis.Aslongasthesecollectionsremainsterile, withclinicallyrelevantandapplicableendpointsisneeded
treatment is generally conservative. However, in one tocomparetheendoscopicandsurgicalstep-upapproach
third of patients infection of necrosis occurs. Infected inordertoreachasoundevidence-basedconclusionabout
necrosis is associated with a mortality rate of around thesuperiorityofeithertreatmentmodality.
30% (range 12-39%) [1,3-5] and is virtually always an
indicationforinvasivetreatment. Methods
The current treatment of choice is a surgical step up Studyobjectives
approach [6]. This approach consists of percutaneous The primary aim of this study is to investigate whether
catheter drainage, if necessary, followed by surgical anendoscopicstep-upapproachwillreducethecombined
necrosectomy. A recent randomised controlled trial primaryendpointofdeathandmajorcomplications,aswell
demonstrated that this approach reduces death and as the secondary endpoints, hospital stay and costs, as
major complications from 69% to 40% compared with compared to a surgical step-up approach in patients
primary open necrosectomy [2]. Furthermore, catheter withinfectednecrotisingpancreatitis.
drainage obviates the need of surgical necrosectomy A secondary aim is to investigate whether endoscopic
and associated complications in 35% of patients [2]. transluminal drainage (ETD) is effective in preventing
Although this trial did not compare minimally invasive necrosectomy.
necrosectomytoopennecrosectomyitdidshowsuperior-
ity of the ‘surgical step-up approach’ to primary open Design
necrosectomy. A promising alternative gaining world- The TENSION trial is a randomised controlled, parallel-
wide popularity is endoscopic transluminal drainage group,superioritymulticentertrial.Patientswillberandomly
andnecrosectomy.Theseprocedurescanbeperformed allocated using an internet-based randomisation program
under procedural sedation, thereby avoiding general (Academic Medical Center) to the endoscopic or surgical
anaesthesia [7]. Furthermore, abdominal wall incision step-upapproach.Patientswith(suspected)infectednecrosis
with its related surgical stress and complications such as are eligible for randomisation. The trial is registered in the
incisional hernia, pancreatic fistula and wound infection, ISRCTNregister(ISRCTN09186711).
areevaded.Theendoscopictechniquecanalsobeapplied
in a step-up fashion, consisting of endoscopic translumi- Participatingcenters
nal drainage (ETD) followed, if necessary, by endoscopic Twenty-four hospitals of the Dutch Pancreatitis Study
transluminalnecrosectomy(ETN).Inrecentyears,several Group (DPSG), including all Dutch university medical
observational cohort studies have been published report- centers,participateintheTENSIONtrialandwillenrolpa-
ing the endoscopic treatment of infected necrosis. A tients (see Appendix for a list of all participating centers).
vanBrunschotetal.BMCGastroenterology2013,13:161 Page3of13
http://www.biomedcentral.com/1471-230X/13/161
Interventions will only take place in centers with sufficient Studypopulation
expertise and after the indication for intervention is sup- Allpatientsadmittedortransferredtooneofthe24partici-
portedbyanonlineexpertpanel. pating hospitals of the DPSG with (suspected) infected ne-
crosis and an indication for intervention will be assessed
for eligibility. Patients (or their legal representatives) that
Primaryendpoint
meetthein-andexclusioncriteriawillbeaskedforwritten
The primary endpoint is a composite of death or major
informedconsent.
complications occurring within 6 months following
randomisation.
Majorcomplicationsaredefinedasnewonset(i.e.not Inclusioncriteria
present 24 hours before randomisation) organ failure (cid:1) (Suspected) infectedpancreaticand/or
(cardiovascular, pulmonary or renal), bleeding requiring extrapancreaticnecrosisandan indication for
intervention, perforation of a visceral organ requiring intervention[2,10](seeTable3fordefinitions).
intervention,enterocutaneousfistularequiringintervention (cid:1) Boththeendoscopicstep-up approachand the
andincisionalhernia(includingburstabdomen)(seeTable1 surgicalstep-up approacharetechnically feasible
fordefinitions). (cid:1) Age≥18years
Secondaryendpoints Exclusioncriteria
(cid:1) theindividual componentsoftheprimaryendpoint (cid:1) Previousintervention(e.g. surgical,endoscopicor
(cid:1) pancreaticocutaneousfistula(seeTable2for percutaneous)forpancreatic necrosis,
definitions) extrapancreaticnecrosisand/orperipancreatic
(cid:1) exocrine orendocrinepancreatic insufficiency collections (seeTable3fordefinitions)
(seeTable2fordefinitions) (cid:1) Acute flareupofchronicpancreatitis
(cid:1) biliarystrictures (cid:1) Indicationforemergencylaparotomybecauseof
(cid:1) wound infections (seeTable2fordefinitions) suspectedabdominalcompartmentsyndrome,
(cid:1) theneedfornecrosectomy (either endoscopically bowelischemia,bleedingorperforationofa
orsurgically) visceralorgan
(cid:1) theneedforadditionalradiological,endoscopicor
surgicalinterventions Randomisation
(cid:1) number ofradiological,endoscopicandsurgical(re-) If a patient with pancreatic and/or extrapancreatic
interventions necrosis shows clinical deterioration and has
(cid:1) totallengthofintensive careandhospitalstay reached the stage to decide on invasive intervention
(cid:1) quality oflife,qualityadjustedlifeyear’s(QALY’s, for (suspected) infected necrosis, the Dutch nation-
withShort Form36and EQ5D) wide expert panel is consulted. This panel, consist-
(cid:1) costsperpatientwithpooroutcomeandcosts ing of 17 experts (9 surgeons, 4 gastroenterologists
perQALY and 4 radiologists) is available 24 hours a day,
(cid:1) totaldirectand indirectmedicalcosts 7 days a week, to assess the indication for interven-
(cid:1) totalnumber ofcross-overbetweengroups tion, the feasibility of both treatment options and
Table1Primaryendpoint:definitions
Event Definition
Organfailure Organfailureisdefinedas:
●Cardiovascular:systolicbloodpressure<90mmHgdespiteadequatefluidresuscitationorneedforvasopressorsupport
●Pulmonary:PaO <60mmHgdespiteFiO 30%,ortheneedformechanicalventilation;
2 2
●Renal:serumcreatinine>177mmol/Lafterrehydrationorneedforhemofiltrationorhemodialysis;
DefinitionsareadaptedfromtheAtlantaclassificationandthesameaspreviouslyusedinthePANTERtrial[2]
Newonsetorganfailure Organfailureoccurringafterrandomisationandnotpresent24hoursbeforerandomisation
Multipleorganfailure Failureof2ormoreorgansystemsonthesameday
Enterocutaneousfistula Enterocutaneousfistulaisdefinedassecretionoffecalmaterialfromapercutaneousdrain,drainagecanalafter
removalofdrains,orfromasurgicalwound,eitherfromsmallorlargebowel;confirmedbyimagingorduringsurgery[2]
Incisionalhernia Incisionalherniaisdefinedasafull-thicknessdiscontinuityoftheabdominalwallandbulgingofabdominalcontents,
withorwithoutobstruction[2]
vanBrunschotetal.BMCGastroenterology2013,13:161 Page4of13
http://www.biomedcentral.com/1471-230X/13/161
Table2Secondaryendpoint:definitions
Event Definition
Pancreaticocutaneousfistula Pancreaticocutaneousfistulaisdefinedasoutput,throughapercutaneousdrain,drainagecanalafterremoval
ofdrains,orfromasurgicalwound,ofanymeasurablevolumeoffluidwithanamylasecontent>3timesthe
serumamylaselevel
Pancreaticinsufficiency ●Exocrineinsufficiencyisdefinedasanabnormalfecalelastasetestortheneedfororalpancreatic-enzyme
supplementationtotreatclinicalsymptomsofsteatorrhea(notpresentbeforeonsetpancreatitis)
●Endocrineinsufficiencyisdefinedasinsulinororalantidiabeticdrugsrequired(notpresentbeforeonsetpancreatitis)
Woundinfection[9] Woundinfectionisdefinedasasuperficialincisionalsurgicalsiteinfection(SSI)andmustmeetthefollowingcriterion:
Infectionoccurswithin30daysaftertheoperativeprocedureandinvolvesonlyskinandsubcutaneous
tissueoftheincisionandthepatienthasatleast1ofthefollowing:
purulentdrainagefromthesuperficial/deepincisionbutnotfromtheorgan/spacecomponentofthesurgicalsite
organismsisolatedfromanasepticallyobtainedcultureoffluidortissuefromthesuperficialincision
atleast1ofthefollowingsignsorsymptomsofinfection:painortenderness,localizedswelling,redness,orheat
thesuperficialincisionisdeliberatelyopenedbysurgeonandisculturepositiveornotcultured.
Aculture-negativefindingdoesnotmeetthiscriterion
anabscessorotherevidenceofinfectioninvolvingthedeepincisionisfoundondirectexamination,
duringreoperation,orbyhistopathologicorradiologicexamination
diagnosisofsuperficial/deepincisionalSSIbythesurgeonorattendingphysician
advises on timing of intervention. In general, inter- Treatmentprotocol
vention is delayed to a phase of the disease at which GroupA(Endoscopictransluminalstep-upapproach)
necrosis is walled-off,usually3–4 weeksafteronset. A Step 1:endoscopic transluminal drainage (ETD) Using
similar expert panel has already proven to be of value procedural sedation, with either i.v. administration of
during the previous Dutch PANTER and PENGUIN midazolamandfentanylorpropofol,endoscopicultrasound
trials[2,10].Afterconsultationoftheexpertpanel,patients guided transluminal drainage of the peripancreatic
eligible for inclusion are randomly assigned to group collection is performed (see Figure 1). Two 7 Fr double
A (endoscopic step-up approach, see Figure 1) or B pigtail stents and a naso-cystic catheter are inserted
(surgical step-up approach, see Figure 2) as shown in the into the collection. The latter will be used for continuous
flowcharts (Figures 3 and 4). Randomisation is performed flushing with 1 liter saline/24 hours. In case of clinical im-
by the study coordinator using an internet-based ran- provement(seecriteriabelow),subsequentnecrosectomyis
domisation program (Academic Medical Center) ensuring avoided. In case a patient does not improve after 72 hours
allocation concealment between groups. Randomisation is and the collection is deemed inadequately drained as ob-
stratifiedaccordingtohospital. served on repeat CECT, additional drainage is performed.
Table3Inclusionandexclusioncriteria:definitions
Event Definition
Pancreaticnecrosis Diffuseorfocalarea(s)ofnon-enhancingpancreaticparenchymaasdetectedoncontrastenhancedCT(CECT)
Extrapancreaticnecrosis PersistentperipancreaticfluidcollectionsonCECTintheabsenceofpancreaticparenchymalnon-enhancement
(Suspected)infectednecrosis ●Infectednecrosisisdefinedasapositivecultureofpancreaticnecrosisorextrapancreaticnecrosisobtained
byfine-needleaspiration(FNA)orthepresenceofgasinthefluidcollectiononCECT.
●Suspectedinfectednecrosisisdefinedaspersistentsepsisorprogressiveclinicaldeteriorationdespitemaximal
supportontheintensivecareunit(ICU)incaseofpancreaticnecrosisorextrapancreaticnecrosis,without
documentationofinfectednecrosisandwithoutothercausesforinfection
Previousintervention Previousexploratorylaparotomyforsuspectedabdominalcompartmentsyndrome,bleedingorsuspected
bowelperforationisonlyallowediftheomentalbursawasnotopened
MODS TheMultipleOrganDysfunctionScore(MODS)rangesfrom0to24,withhigherscoresindicatingmore
severeorgandysfunction
SOFA ScoresontheSequentialOrganFailureAssessment(SOFA)scalerangefrom0to24,withhigherscores
indicatingmoresevereorgandysfunction
vanBrunschotetal.BMCGastroenterology2013,13:161 Page5of13
http://www.biomedcentral.com/1471-230X/13/161
Figure1Endoscopicstep-upapproach.Endoscopicstep-upapproachconsistingofendoscopictransluminaldrainage(ETD)andendoscopic
transluminalnecrosectomy(ETN).Alargeperipancreaticcollectioncontainingfluidandnecrosisisshown.(A)ETD:thecollectionispunctured
throughthegastricwall,followedbyballoondilatationofthetract.Twodouble-pigtailstentsandanasocysticcatheterforcontinuous
postoperativeirrigationareplaced.(B)ETN:thecystostomytractisdilated,thecollectionisenteredwithaendoscope,andnecrosectomyis
performed.(ReprintedfromvanBrunschotetal.[11];copyright2013,withpermissionfromElsevier).
If re-drainage is not indicated (drains are well positioned GroupB(Surgicalstep-upapproach)
in the fluid cavity), clinically unsuccessful or impossible, This approach is similar to the step-up approach used in
thepatientwillproceedtostep2. thePANTERtrial [2,10].
Step 2: endoscopic transluminal necrosectomy (ETN) Step1:percutaneouscatheterdrainage(PCD) A percu-
The cystogastrostomy is dilated up to 18 mm and the taneous 14 to 20 French drain is placed in the peripan-
cavityisenteredwithatherapeuticgastroscopetoper- creatic collection under guidance of CT or ultrasound
form necrosectomy under direct endoscopic vision (see Figure 2). The preferred route is through the left
(see Figure 1). The procedure is completed when most retroperitoneum, thereby facilitating video-assisted retro-
looseadherentnecrotictissueisremoved.Againtwo7 peritoneal débridement (VARD) [13] at a later stage if
Fr double pigtail stents and a naso-cystic catheter for needed. If this is not possible, trans-peritoneal drainage is
continuous lavage will be inserted into the collection. performed. Drains are kept open by flushing with 50 ml
The procedure is repeated in case there is no clinical saline three times daily. In case of clinical improvement,
improvementwithin72hours. thefurthereffectofdrainageisawaited.Ifapatient isnot
vanBrunschotetal.BMCGastroenterology2013,13:161 Page6of13
http://www.biomedcentral.com/1471-230X/13/161
Figure2Surgicalstep-upapproach.Surgicalstep-upapproachconsistingofpercutaneouscatheterdrainage(PCD)andvideo-assisted
retroperitonealdébridement(VARD).(A)Cross-sectionalimageandtorsodepictingaperipancreaticcollection.Thepreferredrouteisthroughthe
leftretroperitonealspacebetweenthekidney,spleenanddescendingcolon.Apercutaneouscatheterdrainisinsertedinthecollectionto
mitigatesepsisandpostponeorevenobviatenecrosectomy.Theareaofdetailisshownin(B).(C)A5cmsubcostalincisionismadeandthe
percutaneousdrainisfollowedintothecollection.Thefirstnecrosisisremovedunderdirectvisionwithalonggraspingforceps,followedby
furtherdebridementundervideoscopicassistance(D).(ReprintedfromvanBrunschotetal.[11];copyright2013,withpermissionfromElsevier).
improvingandacollectionisdeemedinadequatelydrained If drainage technically fails in the endoscopic group a
on repeat CECTafter 72 hours, additional drainage is PCD is placed. In case of clinical or technical failure of
performed.Ifthisisnotpossible,orifa seconddrainageis PCD,surgicalnecrosectomyisperformed.Bothapproaches
clinicallyunsuccessful(seecriteriaforclinicalimprovement are performed, according to a strict protocol, only in
below)thepatientwillproceedtostep2. participating centers with documented expertise and, if
necessary, under supervision of an expert. Sufficient
Step 2: video-assisted retroperitoneal debridement expertise is defined as having performed at least ten
(VARD,ifnotpossiblelaparotomy) VARDisadrain- independent VARD procedures or ten independent endo-
guided, minimally invasive retroperitoneal procedure scopic transluminal drainage procedures and more than
requiring a 5 cm flank incision according to the previ- five endoscopic transluminal necrosectomies. In case of
ously published technique [13,14] (see Figure 2). Using insufficientlocalexperience,thepatientistransferredtoa
the retroperitoneal drain for guidance, the collection is tertiaryreferralcenterwithsufficientexperience.
entered and only loosely adherent necrosis is removed
under video-assistance. At the end of the procedure Criteriaforclinicalimprovement
two large bore surgical drains are inserted. A continuous Criteria similar to the PANTER trial are used to define
post-operative lavage system (building up to 10 litres clinicalimprovement,failureandtodecidetogotothe
saline per 24 hrs) is installed. In case of absence of next step [2,10]. Each step is evaluated 72 hours after
clinical improvement (see criteria below) and repeated intervention and considered successful in case of clinical
CECTshowsfociofpotentiallyinadequatedrainage,VARD improvement.Clinicalimprovementisdefinedas:improved
isrepeated.IfVARDistechnicallynotfeasible,debridement function of at least two organ systems (i.e. circulatory,
bylaparotomyisperformed. pulmonary, or renal) or improvement of two out of three
vanBrunschotetal.BMCGastroenterology2013,13:161 Page7of13
http://www.biomedcentral.com/1471-230X/13/161
Figure3FlowchartTENSIONtrialaccordingtoCONSORT[12].
parametersofinfection(i.e.C-reactiveprotein,leucocytes, Datacollection
or temperature). Clinical failure is defined as the absence Clinical data with regard to baseline characteristics and
ofclinicalimprovementorincaseofclinicaldeterioration. outcome are collected during hospital admission using a
Ifthereis,atevaluationoranymomentthereafter,clinical standardised case record form (CRF). An independent
failure the next step or next necrosectomy is performed. monitor checksall endpointsand atleast10% of theCRF
Deterioration by other infectious causes than infected datawithon-sitesourcedata.
necrosis(e.g.aurinarytractinfection)isexcluded.
Follow-up
Patientsareobservedduringtheirhospitalstay.Outpatient
Generaltreatmentregimen follow-up visits are scheduled at the discretion of the
All patients receive enteral nutrition. If oral feeding is responsible physician, but always 3 and 6 months after
not tolerated or insufficient, a nasojejunal feeding tube randomisation and 3 and 6 months after discharge.
is introduced and enteral feeding is started. If the During these visits all patients will undergo a routine
requiredcaloricintakecannotbereachedviatheenteral CECT, exocrine and endocrine pancreatic function tests
route, the patient will receive (additional) parenteral (i.e. blood glucose measurements and fecal elastase test),
nutrition. All patients with (suspected) infected necrosis andreceiveacombinedquestionnaire(SF-36[15],EQ-5D
will receive broad-spectrum antibiotic therapy according [16],andHealthandLabour[17]).
to institutional protocols. Antibiotic treatment is tailored
basedonbloodculturesandculturefrommaterialcollected Safety
during drainage or surgical procedures. If cultures remain At regular intervals, an independent data safety and moni-
negative, antibiotic treatment is stopped. Before retraction toringcommittee(DSMC)willevaluatetheprogressofthe
orremovalofapercutaneousdrainorthepigtailstentsthe trial and examine the unblinded safety variables [18]. All
remainingcavityisvisualized. physicians involved in the study will repetitively be asked
vanBrunschotetal.BMCGastroenterology2013,13:161 Page8of13
http://www.biomedcentral.com/1471-230X/13/161
Figure4FlowcharttreatmentprotocolTENSIONtrial.
to report any potential adverse events. These events will form prior to randomisation. The TENSION trial is
be listed and discussed with the DSMC. All possible registered in the ISRCTN register with identification
adverse events will also be reported to the Central numberISRCTN09186711.Afterapprovaloftheprotocol,
Committee on Research involving Human Subjects and noamendmentsonstudydesignweremade.
the institutional review board (IRB). Adverse events are
defined as ‘any undesirable experience occurring to a Statisticalaspects
subjectduringaclinicaltrial,whetherornotconsidered Samplesizecalculation
relatedtotheintervention’.Theoutcomeofthemeeting TheTENSION trial is a superiority trial, hypothesizing
of the DSMC will be discussed with the trial steering a reduction in the primary endpoint in favour of the
committeeandsenttotheIRB. endoscopic step-up approach. Combined results of
published non-randomised studies on ETN were used
Ethics tocalculatethesamplesize.These cohort studiesshow
Thestudyisperformedinaccordancewiththedeclaration that ETN results in a combined death and major com-
of Helsinki and the Dutch Medical Research Involving plicationrateof17%[19-24].Thepreviousrandomised
Human Subjects Act. The IRB of the Academic Medical PANTER trial showed a combined death and major
Centre Amsterdam approved the protocol on the 31th complicationrate of40% for thesurgical step-up approach
of January 2011. Secondary approval was sought from [2]. Furthermore,inthesurgical group anincisionalhernia
alllocalethicscommittees.Informedconsentwillbeob- rate of 7% was seen [2]. Assuming that some patients will
tained from each participating patient in oral and written develop an incisional hernia in the surgical group without
vanBrunschotetal.BMCGastroenterology2013,13:161 Page9of13
http://www.biomedcentral.com/1471-230X/13/161
having another primary endpoint, the prevalence of death providers, private household assistance, and indirect
andmajorcomplicationsinthisgroup,includingincisional costs from loss of productivity due to sick leave
hernias is estimated to be 43%. Therefore, an absolute (assessed with the Health and Labour questionnaire).
reduction in primary endpoint of 26% (from 43 to 17%) Total costs per patient are calculated by summing direct
isanticipated.Witha2-sided5%alpha, powerof80%,and medical costs, direct nonmedical costs, and indirect costs
2%losstofollow-up,thesamplesizewassetat98patients. and subsequently compared between groups. Furthermore,
theimpactofdifferencesincomplicationsonthequalityof
Descriptivestatistics life is measured by a generic quality of life questionnaire,
For dichotomous data, frequencies will be presented. the SF-36. In addition to this quality of life questionnaire,
Continuous data will be presented as meanand standard the EQ-5D is completed which screens for the presence
deviation or median and interquartile range. Baseline cri- and severity of problems with mobility, self-care, daily
teria are: age, sex, body mass index, aetiology of pancrea- activities,pain/complaintsandmood.
titis, co-morbidity, American Society of Anaesthesiologist’s
(ASA)classification,CTseverityindex,extentofpancreatic Prematureterminationofthestudy
necrosis, disease severity (SIRS, ICU admission, single No formal interim-analysis is planned. To guarantee
or multiple organ failure), Acute Physiology and Chronic patient’s safety throughout the study, the DSMC will
HealthEvaluation(APACHE)llscore,Imriescore,MODS perform regular safety analyses. When harm (higher
(Table 3), SOFA score (Table 3), C-reactive protein, incidence of SAE’s in one group) occurs, the DSMC
time from onset of symptoms to randomisation,tertiary will discuss potential stopping of the trial prematurely
referral,andconfirmedinfectednecrosis(bacterialculture with the trial steering committee. Since this is the first
offirstintervention). randomised trial comparing both approaches, and hence
all data arising from this trial, regardless of its outcome,
Analyses will influence treatment policy worldwide, the trial will
There will be a blinded outcome assessment after the last notbestoppedforfutility.
patientcompletedfollow-upforallprimaryandsecondary
endpoints. Both intention-to-treat and per-protocol ana- Discussion
lyses will be performed. In intention-to-treat analysis, all The TENSION trial is designed to answer the question
patients are analysed according to their initially assigned whether an endoscopic step-up approach will lead to a
study arm regardless of adherence to study protocol, reduction of death and major complications compared to
whichistheprimaryanalysisofthestudy.Occurrences a surgical step-up approach in patients with (suspected)
of the primary and secondary endpoints are compared infected necrosis. The TENSION trial will also investigate
between treatment groups. Comparison of the primary whetherpancreaticfistula,exocrineorendocrinepancreatic
endpoint will be expressed in terms of a relative risk insufficiency,lengthofICUandhospitalstay,qualityoflife
and corresponding 95% confidence intervals. A two- andcostsarereducedbytheendoscopicstep-upapproach.
tailed P<0.05 is considered statistically significant. Inrecentyears,minimallyinvasiveapproachesaregrad-
Subsequent analyses are directed at secondary end- ually replacing (primary) open necrosectomy. Minimally
points. Predefined subgroup analysis will be performed invasive approachesaim atminimizingsurgical stressand
for patients with and without (multiple) organ failure have proven to reduce complications. In the PANTER
(seeTable1fordefinitions)beforerandomisation,institution trial,thesurgicalstep-upapproachreducedthecombined
and time between onset of symptoms and randomisation deathandmajorcomplicationratefrom69%to40%[2].
(<28or >28days).Aformaltestofinteractioninalogistic- Furthermore, the PANTER trial showed that 35% of
regressionmodelisusedtoassesswhethertreatmenteffects patients with infected necrotising pancreatitis achieve
differ significantly between subgroups. In case of skewed complete recovery after percutaneous drainage only,
randomisation (i.e. statistically significant differences in without the need for surgical debridement. Although
baseline variables), an adjusted analysis will be performed thecombineddeathandmajorcomplicationrateisstill
usingmultivariablelogisticregression. high, the surgical step-up approach should at present
be considered as the current standard of care worldwide
Additionalanalyses [2]. Drainage is based on the hypothesis that alleviating
Directandindirectmedicalandnon-medicalcostsofboth pressureofaninfectedcollectionmayimprovethepatient’s
treatment strategies for the follow-up period of 6 months clinical condition and thereby leaving the necrotic tissue
after randomisation will be compared. All costs will be to be dealt with by the patient’s own immune system.
estimated based on the actual input in terms of resource Endoscopic transluminal drainage can be applied accord-
use (i.e. interventions, diagnostic procedures, hospital ingtothesamerationale.Therefore,wechosetoinstitute
and ICU stay), personnel, medication, visits to healthcare thestep-upapproachinbothstudyarms.Duetothelarge
vanBrunschotetal.BMCGastroenterology2013,13:161 Page10of13
http://www.biomedcentral.com/1471-230X/13/161
differencesintreatmentsbetweenbothgroups,blindingis P. Fockens, MD PhD, Dept. of Gastroenterology and
not feasible. To partially compensate for this, outcome Hepatology, Academic Medical Center, University of
assessmentisblinded. Amsterdam(chair)
In the TENSION trial only patients with (suspected) H.G. Gooszen, MD PhD, Dept. of OR/Evidence Based
infected necrosis are included since the main indication Surgery,RadboudUniversity NijmegenMedical Center
for intervention in necrotising pancreatitis is nowadays M.A.Boermeester,MDPhD,Dept.ofSurgery,Academic
considered to be infected necrosis [25-28]. Patients with MedicalCenter
sterilenecrosiscanoftenbesuccessfullymanagedconser- M.J.Bruno,MDPhD,Dept.ofGastroenterology,Erasmus
vatively(i.e.withoutanyformofintervention)[28-30]. MC,UniversityMedicalCenter
Acompositeendpointofdeathandmajorcomplications C.H.C.Dejong,MDPhD,Dept.ofSurgeryandNUTRIM
was chosen because a study powered to demonstrate a School for Nutrition, Toxicology and Metabolism,
clinically relevant difference in death alone would require MaastrichtUniversityMedicalCenter
an unrealistic large sample size of over 2000 patients. R. Timmer, MD PhD, Dept. of Gastroenterology,
In addition, previous studies have shown that major St Antonius Hospital
complications have high impact in terms of quality of life B.J.M. Witteman, MD PhD, Dept. of Gastroenterology,
inpatientswithnecrotisingpancreatitis[2,8]. GelderseValleiHospital
Apotentiallimitationoftheendoscopicapproachisthat
periproceduralcomplications(e.g.perforationorbleeding)
may be more difficult tomanage when compared to peri- Expert panel
proceduralcomplicationsoccurringduringsurgicalnecro- M.A. Boermeester, MD PhD, Dept. ofSurgery, Academic
sectomy. A systematic review and randomised trial have MedicalCenter,Amsterdam
suggested that endoscopic treatment of infected necrosis T.L.Bollen,MD,Dept.ofRadiology,StAntoniusHospital,
is feasible and associated with lower or comparable com- Nieuwegein
plicationratesthansurgery[7,8].Furthermore,endoscopic M.Bruno,MDPhD,Dept.ofGastroenterology,Erasmus
drainage and necrosectomy are advanced interventions MC,UniversityMedicalCenter,Rotterdam
that not only require the expertise from an interventional V.C. Cappendijk, MD, Dept.of Radiology, Jeroen Bosch
endoscopist, but also the dedicated involvement of inter- Hospital,'s-Hertogenbosch
ventional radiologists and pancreatic surgeons to manage C.H.C.Dejong,MDPhD,Dept.ofSurgeryandNUTRIM
potentialcomplications.Forthisreasontheendoscopicin- School for Nutrition, Toxicology and Metabolism,
terventions in theTENSION trial will only be performed MaastrichtUniversityMedicalCenter,Maastricht
inexpertcenterswithmultidisciplinaryexpertise. C. van Eijck, MD PhD, Dept. of Surgery, Erasmus MC,
UniversityMedicalCenter,Rotterdam
Conclusion P. Fockens, MD PhD, Dept. of Gastroenterology and
TheTENSION trial is a randomised controlled multicen- Hepatology, Academic Medical Center, University of
ter trial designed to show a reduction in the composite Amsterdam,Amsterdam
primary endpoint of death and major complications, as H. van Goor, MD PhD, Dept. Of Surgery, Radboud
well in hospital stay and costs following an endoscopic UniversityNijmegenMedicalCenter,Nijmegen
transluminal step up approach compared with a surgi- H.G. Gooszen, MD PhD, Dept. of OR/Evidence Based
calstepupapproachinpatientswithinfectednecrotis- Surgery, Radboud University Nijmegen Medical Center,
ingpancreatitis. Nijmegen
J.W. Haveman, MD PhD, Dept. of Surgery, University
Appendix MedicalCenterGroningen,Groningen
TENSIONcommitteemembers H.S. Hofker, MD PhD, Dept. of Surgery, University
Steering committee MedicalCenterGroningen,Groningen
S. van Brunschot, MD, Dept. of Gastroenterology and J.S. Laméris, MD PhD, Dept. of Radiology, Academic
Hepatology, Academic Medical Center, University of MedicalCenter,Amsterdam
Amsterdam and dept. of OR/Evidence Based Surgery, K.P.vanLienden,MDPhD,Dept.ofRadiology,Academic
RadboudUniversityNijmegenMedicalCenter MedicalCenter,Amsterdam
H.C.vanSantvoort,MDPhD,Dept.ofSurgery,University V.B. Nieuwenhuijs, MD PhD, Dept. of Surgery, Isala
MedicalCenterUtrecht Clinics,Zwolle
M.G.H.Besselink,MDPhD,Dept.ofSurgery,Academic J.W.Poley,MDPhD,Dept.ofGastroenterology,Erasmus
MedicalCenter MC,UniversityMedicalCenter,Rotterdam
O.J. Bakker, MD PhD, Dept. of Surgery, University A.F.M.Schaapherder,MDPhD,Dept.ofSurgery,Leiden
MedicalCenterUtrecht UniversityMedicalCenter,Leiden