Documents

2010 Prophylaxis Post Ercp Pancreatitis

Categories
Published
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Description
,
Transcript
  European Society of Gastrointestinal Endoscopy (ESGE)Guideline: Prophylaxis of post-ERCP pancreatitis  Authors  J.-M. Dumonceau 1 , A. Andriulli 2 , J. Deviere 3 , A. Mariani 4 , J. Rigaux 3 , T. H. Baron 5 , P. A. Testoni 4 Institutions  Institutions are listed at the end of article. submitted  7 January 2010 accepted after revision 15 March 2010 BibliographyDOI  http://dx.doi.org/10.1055/s-0029-1244208Endoscopy 2010; 42:503 – 515 © Georg ThiemeVerlag KG Stuttgart · New YorkISSN 0013-726X Corresponding author  J.-M. Dumonceau, MD, PhD Division of Gastroenterology and Hepatology Geneva University Hospitalsrue Micheli-du-Crest 241205 GenevaSwitzerlandFax: +41-22-3729366 jmdumonceau@hotmail.com Guidelines  503 1. Introduction ! Endoscopic retrograde cholangiopancreatogra-phy (ERCP) has become almost exclusively a ther-apeutic procedure. Of all the currently performedgastrointestinal endoscopic procedures it carriesthe highest complication rate. Complications of ERCP include pancreatitis, bleeding, cholangitis,cholecystitis, and perforation. Of these complica-tions, post-ERCP pancreatitis (PEP) is the mostfrequent.It ismostoftenclinically mild or moder-ate in severity but in about 10% of cases it is se-vere and potentially fatal. Patient-related andprocedure-related risk factorsfor PEPare well de-fined. Recently, effective measures to prevent PEPhavebeenidentified,andincludeimprovementincannulation techniques as well as pharmacologi-cal and instrumental interventions.The aim of the guideline is to provide a frame-work to caregivers to implement available meth-odstominimizetheincidenceandseverityofPEP.The recommendations are not designed to be ri-gid and cannot replace clinicaljudgment. 2. Methods ! The European Society of Gastrointestinal Endos-copy (ESGE) commissioned this guideline whichwas then endorsed by its Governing Board. Themethod used to formulate the guideline is sum-marized as follows. In 2009 a preliminary litera-ture search was performed by the correspondingauthor. Original papers were identified by asearch of PubMed/MEDLINE, The Cochrane Li-brary, Embase, and the internet, with searchterms  “ ERCP ”  and  “ pancreatitis. ”  Articles werefirst selected by title. Their relevance was thenconfirmed by review of the corresponding ab-stract. Publications in non-English languages andthose whose content was considered irrelevantwere excluded. This initial search focused on fullypublished prospective studies, particularly ran-domized controlled trials (RCTs), though retro-spective analyses and case series were also inclu-ded if they addressed topics not covered in theprospective studies. Additional articles wereidentified by manually searching the referencelists of retrieved papers. A summary of the searchfindings was presented to the ESGE GoverningBoard.The commissioned authors met three times andsubsequently developed the guideline and incor-porated recommendations from the members of the Governing Board. In November 2009, thefinaldraft was sent to all individual ESGE members.After incorporation of comments made by the in-dividualESGEmembers,themanuscriptwasthensent totheEditorialBoardof thejournalEndosco-py for critique and international peer review. Thefinal wording of the guideline document wasagreed upon byall of the authors. Categories of evidence The strength of the evidence used in this guide-line was that recommended by the Scottish Inter-collegiate Guidelines Network [1]. The ratings of levels of evidence are summarized below:Pancreatitis is the most common complication of endoscopic retrograde cholangiopancreatogra-phy (ERCP). Risk factors for post-ERCP pancreati-tis (PEP) are both patient-related and procedure-related. Identification of patients at high risk forPEP is important in order to target prophylacticmeasures. Prevention of PEP includes administra-tion of nonsteroidal inflammatory drugs(NSAIDs), use of specific cannulation techniques,and placement of temporary pancreatic stents.The aim of this guideline commissioned by theEuropean Society of Gastrointestinal Endoscopy(ESGE) istoprovide practical, graded, recommen-dations for the prevention of PEP. Dumonceau J-M. et al. Guidelines for PEP prophylaxis …  Endoscopy 2010; 42: 503 – 515    D  o  w  n   l  o  a   d  e   d   b  y  :   T   h   i  e  m  e   V  e  r   l  a  g  s  g  r  u  p  p  e .   C  o  p  y  r   i  g   h   t  e   d  m  a   t  e  r   i  a   l .  1++  High quality meta-analyses, systematic reviews of RCTs,or RCTs with a very low riskof bias 1+  Well-conducted meta-analyses, systematic reviews of RCTs,or RCTs with a low riskof bias 1 –  Meta-analyses, systematic reviews, or RCTs with a high riskof bias 2++  High quality systematic reviews of case – control or cohortstudies; high quality case – control or cohort studies with avery low riskof confounding, bias, or chance, and a highprobability that the relationship is causal 2+  Well-conducted case – control or cohort studies with a lowriskof confounding, bias, or chance, and a moderateprobability that the relationship is causal 2 –  Case – control or cohort studies with a high riskof confounding, bias, or chance and a significant risk that therelationship is not causal 3  Non-analytic studies, e.g. case reports, case series 4  Expert opinion. Grading of recommendations Recommendationswerebasedonthelevelofevidencepresentedin support and were graded accordingly [1]. This grading is sum-marized below: A  Atleastonemeta-analysis,systematicreview,orRCTratedas1++anddirectlyapplicabletothetarget population;  or  asys-tematic review of RCTs;  or   a bodyof evidence consistingprincipallyof studies rated as 1+ directly applicable to thetarget population and demonstrating overall consistency of results B  A bodyof evidence including studies rated as 2++ directlyapplicable to the target population and demonstratingoverall consistency of results;  or   extrapolated evidence fromstudies rated as 1++ or 1+ C  Abodyofevidenceincludingstudiesratedas2+directlyapplicabletothetargetpopulationanddemonstratingoverallconsistencyofresults; or  extrapolatedevidencefromstudiesratedas2++ D  Evidencelevel 3 or 4;  or   extrapolated evidence from studiesrated as 2+.For interventions analyzed in a singlestudy, no recommendationwas made. 3. Summary of statements and recommendations !  Pancreatitis is the most frequent complication after ERCP withan incidence of 3.5% in unselected patients; it is of mild or moderate severity in approximately 90% of cases. Independent  patient-related and procedure-related risk factors for PEP arelisted in ● Table1 . Risk factors synergistically increase the riskof PEP   (Evidencelevel 1+) .  There is no evidence that hospital ERCP volumehas an influenceon the incidence of PEP; data about a potential relationship be-tween PEP incidence and endoscopist case volume are conflict-ing. Low annual case volumes, of endoscopists and centers, areassociated with higher ERCP failure rates  (Evidence level 2+) .  Serumamylasevalueslessthan1.5timestheupperlimitofnor-mal(ULN),obtainedat2 – 4hourspost-ERCP,almostexcludePEP;valuesmorethan3or5timestheULNat4 – 6hourspost-ERCP haveincreasingpositivepredictivevaluesforPEP   (Evidencelev-el 2+) . It is recommended that serum amylase be determined in patients to be discharged on the day of ERCP; patients withamylase values less than 1.5 times ULN can be discharged with-out concern about risk of PEP   (Recommendation grade B) .  Nonsteroidal anti-inflammatory drugs (NSAIDs) reduce the in-cidence of PEP; effective PEP prophylaxis has only been demon-strated using 100 m  g of diclofenac or indomethacin adminis-tered rectally  (Evidence level 1++) . Routine rectal administra-tion of 100 m  g of diclofenac or indomethacin, immediately be- foreorafterERCP, isrecommended (RecommendationgradeA) .  Nitroglycerin reduces the incidence of PEP; however, when ad-ministered transdermally, it is ineffective  (Evidence level 1++) . Side effects such as transient hypotension and headache mayoccur. We do not recommendthe routine useofnitroglycerinfor  prophylaxis of PEP   (Recommendation grade A) .  Cephtazidime reduced the incidence of PEP in a single study (Evidencelevel 1 – ) . Further data are needed before recom-mending cephtazidime for the prophylaxis of PEP   (Recommen-dation grade C) .  Based on an ad hoc meta-analysis of results from 10 high qual-ity RCTs, somatostatinprovedtobeineffective inpreventingPEP  (Evidencelevel 1++) . We do not recommend universal admin-istration of prophylactic somatostatin in average-risk patientsundergoing ERCP   (Recommendation grade A) . Administrationof somatostatin might be more efficacious using specific doseschedules, but cautionisneededwhen interpretingtheresultsof subgroup analyses astheyoften exaggeratedifferencesbetweentreatments in RCTs.  Octreotide administration did not affect the overall incidence of PEP when data from eight high quality trials were pooled  (Evi-dencelevel 1++) . Prophylaxis with octreotide is not recommen-ded (RecommendationgradeA) .Infuturestudiestheefficacyof  prophylactic administration of octreotide should be evaluatedusing a dose greater than or equal to 0.5mg.  Prophylaxis with gabexate or ulinastatin does not reduce theincidence of PEP   (Evidencelevel 1++) . Neither drug is recom-mended for prophylaxis of PEP   (Recommendation grade A) .  There is no evidence that glucocorticoids, drugs reducing sphincter of Oddi pressure (other than nitroglycerin), antioxi-dants, heparin, interleukin-10, or some anti-inflammatorydrugs (other than diclofenac and indomethacin), such as pen-toxifylline, semapimod and the recombinant platelet-activating  factor acetylhydrolase reduce the incidence of PEP   (Evidencelevels from 1 –  to 1++) . None of these drugs is recommended for PEP prophylaxis  (Recommendation grade A) .  There is no evidence that the incidence of PEP is influenced by patient positionduringERCP  (Evidencelevel2++) .Therefore,norecommendation is made regarding patient position.  Trauma resulting fromrepeatedattemptsat biliarycannulationhas been proven to be a risk factor for the development of PEP  (Evidencelevel 2++) . The number of cannulation attemptsshould be minimized  (Recommendation grade B) .  Injection of contrast medium into the pancreatic duct is an in-dependent predictor of PEP   (Evidence level 1+) . If pancreatic duct injection occurs incidentally or is required, the number of injections and volume of contrast medium injected into the pancreatic duct should be kept as low as possible  (Recommen-dation grade B) .  Compared with traditional, high-osmolality contrast agents,low-osmolality contrast agents are costlier but are not asso-ciatedwithreductionintheratesofPEP  (Evidencelevel1 – ) .Theroutine use of these agents for ERCP is not recommended  (Re-commendation grade B) . Guidelines 504 Dumonceau J-M. et al. Guidelines for PEP prophylaxis …  Endoscopy 2010; 42: 503 – 515    D  o  w  n   l  o  a   d  e   d   b  y  :   T   h   i  e  m  e   V  e  r   l  a  g  s  g  r  u  p  p  e .   C  o  p  y  r   i  g   h   t  e   d  m  a   t  e  r   i  a   l .   Use of carbon dioxide (CO2) as a replacement for air for luminalinsufflation duringERCP does not influence the incidence of PEP but decreases the incidence and severity of post-procedural ab-dominal pain  (Evidencelevel 1+) . Carbon dioxide is recommen-ded for insufflation, and might be particularly useful for outpa-tient ERCPs, to reduce post-procedural abdominal pain and toavoid confusion with PEP   (Recommendation grade B) .  Fordeep biliarycannulation, the wire-guided technique reducesthe risk of PEP and increases the success rate of primary cannu-lation when compared with the standard contrast-assistedmethod  (Evidencelevel 1++) . The wire-guided technique is re-commended for deep biliary cannulation  (Recommendationgrade A) .  The incidence of post-sphincterotomy pancreatitis is not influ-encedbythetypeofelectrosurgicalcurrent used(whether pure-cut or blended)  (Evidencelevel 1+) . Blended current is recom-mended for biliary sphincterotomy, particularly in patients at high risk of bleeding   (Recommendation grade A) .  Data about the usefulness and safety of pancreatic guide wire placement to facilitate biliary cannulation in difficult cases areconflicting. Prophylactic pancreatic stent placement decreasesthe incidence of PEP with this technique  (Evidencelevel 2+) .Pancreatic guide wire assistance may facilitate biliary cannu-lation mostly in the case of inadvertent but repeated cannula-tion of the pancreatic duct; if this method is used, a pancreatic stent should be placed for PEP prophylaxis  (Recommendationgrade B) .  Various techniques of precut biliary sphincterotomy have beendescribed; the fistulotomy technique may present a lower inci-dence of PEP than standard needle-knife sphincterotomy, but  further RCTs are required to determine which technique is safer and more effective, based upon the papillary anatomy. There isno evidence that the success and complication rates of biliary precut are affected by the level of endoscopist experience in thistechnique but published data only report on the experience of one endoscopist   (Evidence level 2 – ) . Prolonged cannulation at-tempts using standard techniques may impart a risk for PEP  greater than the precut sphincterotomy itself   (Evidencelevel 2+) . Precut sphincterotomy should be performed by endoscopistswith expertise in standard cannulation techniques  (Recom-mendation grade D) . The decision to perform precut biliarysphincterotomy, the timing, and the technique are based onanatomic findings, endoscopist preference and procedural indi-cation  (Recommendation grade C) .  Compared with endoscopic sphincterotomy, endoscopic papil-lary balloon dilation (EPBD) using small-caliber balloons ( ≤ 10mm) is associated with a significantly higher incidence of PEP and significantly less bleeding   (Evidencelevel 1++) . EPBD is not recommended as an alternative to sphincterotomy in routineERCP but may be useful in patients with coagulopathy and al-tered anatomy (e.g. Billroth II)  (Recommendation grade A) . If balloon dilation is performed in young patients, the placement of a prophylactic pancreatic stent should be strongly considered (Evidencelevel 4; Recommendation grade D) .  Potentialadvantagesofperforminglarge-balloondilationinad-ditiontoendoscopicsphincterotomyforextractionofdifficult biliarystonesremainunclear   (Evidencelevel3) .Endoscopic sphincterotomypluslarge-balloondilationdoesnotseemtoin-creasetheriskofPEPandcanavoidtheneedformechanicalli-thotripsyinselectedpatients,butnotenoughdataareavailabletorecommendroutineuseoverbiliarysphincterotomyaloneinconjunctiontolithotripsytechniques (RecommendationgradeD).  In patients undergoing pancreatic sphincter of Oddi manome-try, use of the standard perfusion catheter, without an aspira- Adjusted odds ratios (95%CIin parentheses except whereindicated otherwise)Pooled incidence of PEPin patients with vs. thosewithout risk factorTable1  Independent riskfactors for post-ERCP pancrea-titis (PEP).* Patient-related risk factors Definite risk factorsSuspected SOD 4.09 (3.37 – 4.96) 10.3% vs. 3.9%Female gender 2.23 (1.75 – 2.84) 4.0% vs. 2.1%Previouspancreatitis 2.46 (1.93 – 3.12) 6.7% vs. 3.8%Likely risk factorsYounger age 1.09 – 2.87 (range 1.09 – 6.68) 6.1% vs. 2.4%Non-dilated extrahepatic bile ducts NR 6.5% vs. 6.7%Absence ofchronic pancreatitis 1.87 (1.00 – 3.48) 4.0% vs. 3.1%Normal serum bilirubin 1.89 (1.22 – 2.93) 10.0% vs. 4.2% Procedure-related risk factors Definite risk factorsPrecut sphincterotomy 2.71 (2.02 – 3.63) 5.3% vs. 3.1%Pancreatic injection 2.2 (1.60 – 3.01) 3.3% vs. 1.7%Likely risk factorsHigh number ofcannulation attempts †  2.40 – 3.41 (range 1.07 – 5.67) 3.7% vs. 2.3%Pancreatic sphincterotomy 3.07 (1.64 – 5.75) 2.6% vs. 2.3%Biliary balloon sphincter dilation 4.51 (1.51 – 13.46) 9.3% vs. 1.9%Failure to clear bile duct stones 3.35 (1.33 – 9.10) 1.7% vs. 1.6% ERCP,endoscopicretrogradecholangiopancreatography;CI,confidenceinterval;SOD,sphincterofOddidysfunction;NR,notreported.* For definite risk factors, adjusted odds ratios and pooled incidences of PEP are reproduced from Masci et al. [10]. For likely riskfactors,adjustedoddsratiosarereproducedfromincludedstudiesthatidentifiedthecharacteristicasanindependentriskfactor, while pooled incidences were calculated using figures available in all of the included studies that provided sufficientdata for calculation (see text for details about included studies) [2,11 – 14]. † “ High ”  (vs. low) number ofcannulation attempts was defined as number of attempts before final cannulation of the desiredduct, and was > 5 or > 1, depending on the studies. Guidelines  505 Dumonceau J-M. et al. Guidelines for PEP prophylaxis …  Endoscopy 2010; 42: 503 – 515    D  o  w  n   l  o  a   d  e   d   b  y  :   T   h   i  e  m  e   V  e  r   l  a  g  s  g  r  u  p  p  e .   C  o  p  y  r   i  g   h   t  e   d  m  a   t  e  r   i  a   l .  tion port, has been shown to increase the risk of PEP comparedwith modified water perfusion catheters  (Evidencelevel 2++) .Pancreatic sphincter of Oddi manometry should be done usingamodified triple-lumen perfusion catheter with simultaneousaspiration or a microtransducer catheter (non-water-perfused) (Recommendation grade B) .  Prophylactic pancreatic stent placement is recommended to prevent PEP in patients who are at high risk for development of PEP. Short 5-Fr diameter plastic pancreatic stents are currentlyrecommended. Passage of the stent from the pancreatic duct should be evaluated within 5 to 10 days of placement and re-tained stents should be promptly removed endoscopically (Evidencelevel 1+; Recommendation grade A). 4. Definitions ! The consensus definition of ERCP complications as proposed byCotton et al. has allowed standardized reporting of the incidenceand severity of PEP [2]. PEP was srcinally defined as  “ clinicalpancreatitis with amylase at least three times normal at morethan 24 hours after the procedure, requiring hospital admissionor a prolongation of planned admission. ”  Some variations existacross studies in the interpretation of   “ clinical pancreatitis, ”  andthis has been defined by some as  “ new or worsened abdominalpain ”  [2],  “ typical pain and symptoms ”  [3], or  “ abdominal painand tenderness ”  [4]. The definition used by Freeman et al. (newor worsened abdominal pain) takes into account patients whoundergo ERCP in the setting of acute pancreatitis or a flare of chronic pancreatitis [2]. The current grading system for the se-verity of PEP is mainly based on the length of hospitalization:mild PEP is defined as need for hospital admission or prolonga-tion of planned admission up to 3 days; moderate PEP is definedby need for hospitalization of 4 – 10 days, and severe PEP by hos-pitalization for more than 10 days, or hemorrhagic pancreatitis,phlegmon (now referred to as pancreatic necrosis), or pseudo-cyst, or need for percutaneous drainage or surgical intervention[5].Althoughthecurrentclassificationsystemallowstheseverityof pancreatitis to be determined in retrospective studies, we re-commend that more specific grading systems of pancreatitis se-verity (e.g. the Atlanta Classification System) be used in futureprospective studies [6].Intheabsenceofchronicpancreatitis,anelevatedserumamylaseis frequently seen 24 hours after ERCP (53% in a prospectivestudy). Abdominal pain in the absence of PEP occurred in 62% of cases in an RCT when air, rather than carbon dioxide, was usedfor luminal insufflation during ERCP [7,8]. Therefore, a standardthreshold level for serum amylase (three times the upper limit of normal [ULN] values 24 hours post-ERCP) and clinical examina-tion of patients by an evaluator blinded to the allocated treat-ment group are important in RCTs that assess the effectivenessof interventions to prevent PEP. 5. Incidence, risk factors, and severity of PEP !  Pancreatitis is the most frequent complication after ERCP withan incidence of 3.5% in unselected patients; it is of mild or moderate severity in approximately 90% of cases. Independent  patient-related and procedure-related risk factors for PEP arelisted in ● Table1 . Risk factors synergistically increase the riskof PEP   (Evidencelevel 1+).Based on a systematic review of 21 prospective studies involvingmore than 16 000 patients [9], PEP was found to be the most fre-quent complication following ERCP with an incidence of 3.47%(95% confidence interval [CI], 3.19% – 3.75%). As defined pre-viously, PEP can be mild, moderate, or severe. Based upon datafrom studies that have included unselected patients, PEP is mildin 45%, moderate in 44%, and severe in 11% of cases, and causesdeath in 3% of cases (95%CI, 1.65% – 4.51%). Stratification of pa-tients into low-risk or high-risk categories for PEP is importantin order to provide adequate pre-procedure information to thepatient and indeciding when toconsider patient referral toater-tiary center.Based on a large meta-analysis [10], three patient-related andtwo procedure-related characteristics are considered definite in-dependent risk factors for PEP ( ● Table1 ). Known or suspectedsphincter of Oddi dysfunction (SOD) presents the strongest asso-ciation, with an incidenceof PEP closeto10%. As onlyfive poten-tial risk factors for PEP were analyzed in that meta-analysis, wealso reviewed prospective, multicenter studies that analyzed po-tential risk factors for PEP using multivariate analysis. Five stud-ies were selected that involved 13 745 patients in total [2,11 – 14]. Patient-related and procedure-related characteristics inde-pendently associated with PEP in at least one of these studiesare reported as likely risk factors in ● Table1  ( “ pancreatic injec-tion ”  corresponded to  ≥ 1 injection and, depending on studies, a “ highnumberofcannulationattempts ” tomorethanfiveormorethan one attempts before cannulation of the desired ducts). Therisk factors presented in ● Table1  are not exhaustive becausenot all potential risk factors have been analyzed. For example,ampullectomy is generally considered to be a definitive risk fac-tor for PEP on the basis of several small prospective studies[15,16].As risk factors for PEP were shown to be independent by multi-variate analysis, they might have a cumulative effect. Freeman etal. calculated the adjusted odds ratio (OR) for various combina-tions of risk factors by using data prospectively collected fromabout 2000 ERCPs: thehighest riskof PEP (42%) wasfound for fe-male patients withanormalserumbilirubin, suspectedSOD,anddifficult biliary cannulation [11]. The actual incidence and sever-ity of PEP in high-risk conditions is estimated using data fromcontrol arms of RCTs in which the effectiveness of prophylacticpancreatic stent placement was evaluated (patients were selec-ted for inclusion based on the presence of SOD, a common bileduct diameter <10mm, precutting, difficult cannulation, sphinc-ter of Oddi manometry, ampullectomy, and also simple endo-scopic sphincterotomy) [15,17 – 19]. Meta-analysis of the controlarms of four such trials found a PEP incidence of 24.1%; 84.4% of cases were mild/moderate and 15.6% were severe [20].  There is no evidence that hospital ERCP volumehas an influenceon the incidence of PEP; data about a potential relationship be-tween PEP incidence and endoscopist case volume are conflict-ing. Low annual case volumes, of endoscopists and centers, areassociated with higher ERCP failure rates  (Evidence level 2+) . Factors that may affect the outcome of ERCP that are specificallyrelated to hospital procedure volume include availability of equipment and adequacy of anesthesia, endoscopic and radiolo-gic support, and nursing assistance. The number of ERCPs per-formed in many centers is not as high as commonly believed: inthree large (regional or national) studies, the median annualnumber of ERCPs was between 49 and 235 [21 – 23]. In one largestudy, the median annual number of ERCPs per endoscopist was Guidelines 506 Dumonceau J-M. et al. Guidelines for PEP prophylaxis …  Endoscopy 2010; 42: 503 – 515    D  o  w  n   l  o  a   d  e   d   b  y  :   T   h   i  e  m  e   V  e  r   l  a  g  s  g  r  u  p  p  e .   C  o  p  y  r   i  g   h   t  e   d  m  a   t  e  r   i  a   l .
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks