()+%&+'*+)'"&+*+"&$,) )0'%"&$ &+)
$*"&#"&"-)*"+0'*("+$
'+')$)' )%%"& $"&"$*)!
&"-)*"+0'$*"&#"
$*"&#""&$&
'()*&+."+!+!()%"**"'&'+!,$+0'""&' +!&"-)*"+0'$*"&#"')(,$"/%"&+"'&"&+,)''%
)+%&&*+"+,+$*"&#"'&()"$+'$'#
$*"&#"
SUPERVISORS
Adjunct Professor Leena Kylänpää
Department of Gastrointestinal Surgery, Abdominal Centre Helsinki University Hospital
University of Helsinki Helsinki, Finland
Adjunct Professor Marianne Udd
Department of Gastrointestinal Surgery, Abdominal Centre Helsinki University Hospital
University of Helsinki Helsinki, Finland
REVIEWERS
Adjunct Professor Markku Heikkinen Department of Internal Medicine Kuopio University Hospital University of Eastern Finland Kuopio, Finland
Antti Siiki, MD, PhD
Department of Gastroenterology and Alimentary Tract Surgery Tampere University Hospital
University of Tampere Tampere, Finland
OPPONENT
Professor Tuomo Rantanen Department of Surgery Kuopio University Hospital University of Eastern Finland Kuopio, Finland
The Faculty of Medicine uses the Urkund system (plagiarism recognition) to examine all doctoral dissertations.
ISBN 978-951-51-7908-1 (paperback) ISBN 978-951-51-7909-8 (PDF) http://ethesis.helsinki.fi Unigrafia
To my family
TABLE OF CONTENTS ... 4
ABSTRACT ... 6
TIIVISTELMÄ ... 8
LIST OF ORIGINAL PUBLICATIONS ... 10
ABBREVIATIONS ... 11
1 INTRODUCTION ... 12
2 REVIEW OF THE LITERATURE ... 14
2.1 ENDOSCOPIC RETROGRADE CHOLANGIOPANCREATOGRAPHY (ERCP) ... 14
2.1.1 Indications for ERCP ... 14
2.1.2 ERCP performance and success ... 16
2.1.3 Biliary cannulation techniques ... 18
2.1.4 Difficult biliary cannulation ... 18
2.1.4.1 Precut techniques ... 20
2.1.4.2 Double guidewire technique (DGW) ... 20
2.1.4.3 Transpancreatic biliary sphincterotomy (TPBS) ... 22
2.1.4.4 Other methods and cannulation algorithm ... 23
2.2 COMPLICATIONS IN ERCP ... 26
2.2.1 Post-endoscopic retrograde cholangiopancreatography pancreatitis (PEP) ... 27
2.2.1.1 PEP prophylaxis ... 30
2.2.2 Cholangitis ... 33
2.2.3 Bleeding ... 33
2.2.4 Perforation ... 34
2.2.5 Other complications ... 34
2.3 PRIMARY SCLEROSING CHOLANGITIS (PSC) ... 35
2.3.1 PSC and ERCP ... 37
2.3.2 Risk factors for post-ERCP complications in PSC ... 38
2.4 DRUG-INDUCED PANCREATITIS ... 39
3 AIMS OF THE STUDY ... 42
4 MATERIALS AND METHODS ... 43
4.1 PATIENT DATA AND CHARACTERISTICS ... 43
4.1.1 Study setting, population, and timing ... 43
4.1.2 Ethical aspects ... 43
4.2 ERCP PERFORMANCE ... 44
4.2.1 Difficult ERCP ... 44
4.2.2 Classification of ERCP complications ... 45
4.3 PSC CLASSIFICATION AND ADVANCED PSC ... 45
4.4 STATISTICAL METHODS ... 46
5 RESULTS ... 47
5.1 STUDY I ... 47
5.2 STUDY II ... 49
5.3 STUDY III ... 51
5.4 STUDY IV ... 53
6 DISCUSSION ... 54
6.1 STUDIES I AND II ... 54
6.2 STUDIES III AND IV ... 56
6.3 FUTURE ASPECTS ... 57
7 CONCLUSIONS ... 59
REFERENCES ... 60
ACKNOWLEDGEMENTS ... 76
ORIGINAL PUBLICATIONS ... 78
Endoscopic retrograde cholangiopancreatography (ERCP) is a key method to explore and treat diseases of the biliary and pancreatic ducts. ERCP facilitates treatment of bile duct stones, bile duct strictures (benign or malign), primary sclerosing cholangitis (PSC) -related strictures and ductal sampling, and bile leakage. ERCP indications in the pancreatic duct include chronic pancreatitis, strictures, pseudocysts, and fistulas. ERCP enables samples to be taken in suspected malignancies. With cholangioscopy or pancreatoscopy, it is possible also to visualize the ducts and take targeted biopsies.
In the ERCP procedure, a flexible duodenoscope is introduced to the duodenum and the papilla major or minor is cannulated, normally with a guidewire, to direct the specimens to the duct explored. ERCP is an invasive procedure with higher complication rates than other endoscopic procedures.
The most common complication is post-endoscopic retrograde cholangiography pancreatitis (PEP) with an incidence of 3-10%. Several risk factors for PEP have been established. Difficult biliary cannulation increases the risk of PEP. In a difficult cannulation situation, advanced cannulation methods, e.g. transpancreatic biliary sphincterotomy (TPBS) and double guidewire technique (DGW), can be performed. The long-term outcomes of these advanced techniques are unknown. Rectal non-steroidal anti- inflammatory drugs (NSAIDs) are recommended as prophylaxis for PEP in the ERCP guidelines.
PSC patients undergo ERCPs when suffering from jaundice or for detection of biliary dysplasia or when cholangiocarcinoma (CCA), which is highly associated with PSC, is suspected. The risk of CCA in PSC is almost 400-fold higher than in the general population. However, the majority of ERCP procedures in PSC are performed as a follow-up, thus, the risk of PEP and other procedure-related complications should be diminished. The role of NSAIDs in PEP risk of PSC patients is ambiguous. PSC is highly associated with inflammatory bowel diseases and autoimmune hepatitis, and these patients use immunosuppressive thiopurines as a maintenance therapy.
Thiopurines are related to increased risk of acute pancreatitis, but their role in acute PEP is unknown.
This thesis includes two studies on biliary cannulation techniques and two studies on PSC patients and the risk of PEP related to medical treatment.
Patients were referred to Helsinki University Hospital (HUS), Meilahti Abdominal Centre Endoscopy Unit between 2007 and 2018. Study II also included patients from other Scandinavian hospitals. In Study I, historical data were collected retrospectively, and 104 patients were interviewed by phone as prospective cut point to determine the long-term outcome after TPBS. Ninety-nine patients were recruited as a control group. Only a few
studies have been published on long-term follow-up in TPBS. According to our results, TPBS appears to be a safe procedure, with no difference in acute or long-term complications compared with the control group. Study II included TPBS and DGW patients in a multicentre randomized prospective investigation by the Scandinavian Association for Digestive Endoscopy group. The result revealed no difference in PEP rate between the two advanced methods, and TPBS seemed to be more effective in achieving deep biliary cannulation. Study III included a retrospective cohort of 2000 ERCPs performed on 931 PSC patients: 1000 ERCPs with rectal administration of 100 mg diclofenac prior to the ERCP, and 1000 ERCPs as a control group. No effect of rectal diclofenac on PEP rate was found. This study was the largest thus far concerning PSC patients and PEP prophylaxis with rectal NSAIDs. In retrospective Study IV, PSC patients were divided into two groups with propensity score matching: the study group (n=177) comprising patients with immunosuppressive thiopurines (azathioprine or mercaptopurine) and the control group containing similar patients (n=177) without thiopurines. No difference in PEP rate was found when comparing the two groups.
Thiopurine use did not increase PEP risk.
In conclusion, in difficult biliary cannulation TPBS is safe method with similar PEP rates to DGW. In PSC, rectal diclofenac seems to be ineffective in PEP prevention; however, immunosuppressive thiopurines do not increase the risk of PEP.
Sappi- ja haimatiehyen tähystystoimenpiteellä (ERCP) tutkitaan ja hoidetaan näiden kohde-elinten sairauksia tai oireita. Toimenpiteen syynä ovat mm.
sappitiehytkivet, bakteerien aiheuttama sappitietulehdus, sappitiehyen kasvaimet ja ahtaumat, sappirakon leikkauksen jälkeinen sappivuoto, sekä sappiteiden harvinainen ahtauttava sairaus, primaari sklerosoiva sappitietulehdus (PSC). Krooniseen haimatulehdukseen liittyvät oireet ovat pääosin haimatiehyen tähystyksen aiheena. ERCP:n tavallisin komplikaatio on haimatulehdus, jonka tarkka tautimekanismi on epäselvä.
Haimatulehdusriski vaihtelee eri tutkimusten mukaan 3–10 %.
Haimatulehduksen riskiä lisäävät mm. potilaan nuori ikä ja naissukupuoli sekä papillan, eli sappi- ja haimatiehyen suun, kanyloinnin aiheuttama mekaaninen trauma. Haimatulehdusriskiä voidaan yrittää pienentää ennen ERCP:tä annetulla tulehduskipulääke -peräpuikolla. Äkillisen haimatulehduksen oireena on ylävatsakipu ja veren amylaasipitoisuuden nousu. Potilas tarvitsee usein ainakin lyhytaikaista sairaalahoitoa ja toipuminen toimenpiteestä pitkittyy.
Noin 25 %:ssa ERCP-toimenpiteistä sappiteiden kanylaatio määritellään vaikeaksi, eli se kestää yli 5 minuuttia, siihen vaaditaan yli viisi yritystä, tai ohjainvaijeri päätyy haimatiehyeen vähintään kaksi kertaa. Tällöin joudutaan käyttämään erilaisia lisätoimenpiteitä, jolloin haimatulehdusriski kasvaa.
Sappi- tai haimatiehyen suun avaamisen, eli sfinkterotomian harvinainen myöhäisvaiheen komplikaatio, tiehyen suulle kehittyvä ahtauma, voi aiheuttaa toistuvia sappitie- tai haimatulehduksia.
Primaari sklerosoiva sappitietulehdus eli PSC on sappiteiden autoimmuunisairaus, joka aiheuttaa vaihtelevia sappiteiden ahtaumia ja voi johtaa maksakirroosiin. Lisäksi sappitiesyövän riski on kohonnut lähes 400- kertaisesti verrattuna muuhun väestöön. PSC-taudin oireiden hoitoon ja sappitiesyövän esiasteiden tutkimiseen ja seurantaan käytetään ERCP- toimenpidettä. PSC on ERCP:n jälkeisen haimatulehduksen yksi riskitekijä.
PSC-potilaista osalla on käytössä haimatulehdusriskiä lisäävä tiopuriini- lääkitys, jonka vaikutusta ERCP:n jälkeiseen haimatulehdusriskiin ei tunneta.
Tämä väitöskirjatyö koostuu kahdesta sappiteiden kanylaatiota käsittelevästä osatyöstä ja kahdesta PSC potilaiden haimatulehdusriskiin liittyvästä osatyöstä.
Osatyö I Haimasfinkterotomian pitkäaikaisvaikutukset seurannassa.
Tutkimuksessa kerättiin potilastietojärjestelmästä takautuvasti 104 haimasfinkterotomiaan päätyneen potilaan tiedot ja verrokkeina toimivat ikä- ja sukupuolikaltaistetut potilaat, joille ei tehty haimasfinkterotomiaa.
Toteutimme puhelinhaastattelututkimuksen 4–10 vuoden seuranta-ajalla.
Tutkimukseen osallistui yhteensä 283 potilasta. Tutkimuksessa selvitettiin
mm. sairastetut haimatulehdukset, haiman vajaatoimintaan liittyvä oireilu ja löydökset, vatsakipuoireilu ja mahdolliset vatsan alueen toimenpiteet.
Havaitsimme, että pitkäaikaisseurannassa haimasfinkterotomia -potilailla ei ollut enempää oireita, vatsakipua tai kroonista haimatulehdusta, eikä myöskään viitettä pitkäaikaisesta haimatiehyen suun arpeutumisesta.
Aiheesta on aikaisemmin julkaistu vain vähän tutkimuksia. Voimme osoittaa, että tarvittaessa tehtävä haimasfinkterotomia on osaavissa käsissä turvallinen toimenpide, jossa pitkäaikaishaittoja ei näyttäisi olevan enempää kuin verrokkiryhmässäkään.
Osatyössä II selvitimme satunnaistetussa prospektiivisessa monikeskus- tutkimuksessa haimasfinkterotomian (n=104) ja kaksoisvaijeritekniikan (n=99) haimatulehdusriskiä ja toimenpiteen onnistumista vaikeassa sappitien kanylaatiossa. Tutkimus tehtiin yhteistyössä seitsemän pohjoismaisen tähystysyksikön kanssa. Havaitsimme, että vaikeassa kanylaatiossa haimatulehdusriski näillä kahdella menetelmällä on samanlainen, mutta toimenpiteessä onnistuminen on yleisempää haimasfinkterotomian avulla. Löydöksillä on todennäköisesti vaikutusta jatkossa hoitosuosituksiin vaikeassa sappiteiden kanylaatiossa.
Osatyössä III selvitimme peräpuikkona ennen ERCP:tä annetun tulehduskipulääkkeen, diklofenaakin, vaikutusta ERCP:n jälkeiseen haimatulehdukseen PSC-potilailla. Keräsimme 1000 toimenpidettä, joissa potilas ei ollut saanut diklofenaakkia ja 1000 toimenpidettä, joissa diklofenaakki oli annettu. Tutkimuksessa vertailtiin haimatulehduksen esiintyvyyttä ja muita komplikaatioita. Havaitsimme, että korkean haimatulehdusriskin PSC-potilailla ei voida tässä aineistossa osoittaa, että diklofenaakki ennaltaehkäisisi haimatulehduksia. PSC-potilailla ei vastaavan laajuista tutkimusta ole aiemmin tehty. Aihe vaatii kuitenkin jatkotutkimuksia, koska tulos on ristiriidassa aiempien satunnaistettujen tutkimusten kanssa.
Osatyössä IV selvitimme PSC-potilaiden tiopuriini-lääkityksen vaikutusta ERCP:n jälkeiseen haimatulehdusriskiin. Lääkitystä käyttäneet potilaat (n=177) kaltaistettiin verrokkien (n=177) kanssa ja selvitimme tiopuriinien vaikutusta ERCP:n komplikaatioihin, erityisesti haimatulehdukseen. PSC- potilaiden liitännäissairauksien (tulehduksellinen suolistosairaus tai autoimmuunimaksatulehdus) hoitoon käytetty immunosuppressiivinen tiopuriini -lääkitys ei näyttäisi lisäävän haimatulehdusriskiä. Tuloksella on merkitystä arvioitaessa ERCP:n turvallisuutta ja eri tekijöiden vaikutusta toimenpidekomplikaatioihin. Aiheesta ei ole aikaisemmin julkaistu laajoja potilastutkimuksia.
Yhteenvetona väitöskirjatyön tuloksista voidaan todeta, että vaikeassa sappitiekanylaatiossa haimasfinkterotomia on turvallinen pitkäaikaisseurannan perusteella, ja sen haimatulehdusriski on samaa luokkaa kuin kaksoisvaijeritekniikassa. PSC potilailla diklofenaakki- peräpuikosta ei ollut hyötyä haimatulehduksen ehkäisyssä, ja toisaalta tiopuriini-lääkityksen ei havaittu lisäävän haimatulehdusriskiä.
This thesis consists of the following publications, referred to in the text by their Roman numerals:
I Koskensalo V, Udd M, Rainio M, Halttunen J, Sipilä M, Lindström O, Kylänpää L. Transpancreatic biliary sphincterotomy for biliary access is safe also on a long-term scale. Surg Endosc. 2021 Jan;35(1):104-112. doi:
10.1007/s00464-019-07364-y. Epub 2020 Jan 28. PMID:
31993817; PMCID: PMC7746558.
II Kylänpää L, Koskensalo V, Saarela A, Ejstrud P, Udd M, Lindström O, Rainio M, Tenca A, Halttunen J, Qvigstad G, Arnelo U, Fagerström N, Hauge T, Aabakken L, Grönroos J.
Transpancreatic biliary sphincterotomy versus double guidewire in difficult biliary cannulation: a randomized controlled trial.
Endoscopy. 2021 Oct;53(10):1011-1019. doi: 10.1055/a-1327- 2025. Epub 2021 Jan 13. PMID: 33440441.
III Koskensalo V, Tenca A, Udd M, Lindström O, Rainio M, Jokelainen K, Kylänpää L, Färkkilä M. Diclofenac does not reduce the risk of acute pancreatitis in patients with primary sclerosing cholangitis after endoscopic retrograde cholangiography. United European Gastroenterol J. 2020 May;8(4):462-471. doi: 10.1177/2050640620912608. Epub 2020 Mar 8. PMID: 32213036; PMCID: PMC7226698.
IV Koskensalo V, Aronen P, Färkkilä M, Kylänpää L, Lindström O, Rainio M, Udd M, Jokelainen K, Tenca A. Use of thiopurines is not a risk factor for post-ERC pancreatitis in patients with primary sclerosing cholangitis. Dig Liver Dis. 2021 Aug;53(8):1020-1027. doi: 10.1016/j.dld.2021.05.009. Epub 2021 Jun 8. PMID: 34116970.
The original publications are reproduced with the permission of their copyright holders.
AIH autoimmune hepatitis
AP acute pancreatitis
ASGE American Society for Gastrointestinal Endoscopy AZA azathioprine
BS biliary sphincterotomy
CBD common bile duct
CCA cholangiocarcinoma
CD Crohn`s disease
CI confidence interval
CT computer tomography
DG diclofenac group
DGW double guidewire technique
ERCP endoscopic retrograde cholangiopancreatography ESGE European Society of Gastrointestinal Endoscopy EUS endoscopic ultrasound
HUS Helsinki University Hospital IBD inflammatory bowel disease
IBDU inflammatory bowel disease, unclassified MRCP magnetic resonance cholangiopancreatography NGS next-generation sequencing
NK needle knife
NNT number needed to treat
NSAID non-steroidal anti-inflammatory drug
OR odds ratio
PD pancreatic duct
PEC post-endoscopic retrograde cholangiography cholangitis PEP post-endoscopic retrograde cholangiopancreatography
pancreatitis
PPS prophylactic pancreatic stent PSC primary sclerosing cholangitis
PTBD percutaneous transhepatic biliary drainage
RR risk ratio
SADE Scandinavian Association for Digestive Endoscopy
SD standard deviation
SOD sphincter of Oddi dyskinesia
TPBS transpancreatic biliary sphincterotomy
UC ulcerative colitis
UEG United European Gastroenterology ULN upper limit of normal
WGC wire-guided cannulation 6-MP 6-mercaptopurin
Endoscopic retrograde cholangiopancreatography (ERCP) is a minimally invasive method to explore and treat hepatobiliary and pancreatic disorders.
Approximately 4000-5000 ERCPs per year are performed in Finland, and, of these, approximately 1500 ERCPs in Helsinki University Hospital (HUS).
The ERCP procedure is performed with a flexible side-viewing duodenoscope. The patient is usually under conscious sedation or in general anaesthesia during the procedure (Sidhu et al. 2019). In ERCP, the imaging requires fluoroscopy to visualize the biliary tree and/or the pancreatic duct with the injected contrast agent. Newer methods, direct duct visualization (cholangioscopy or pancreatoscopy), and endoscopic ultrasound (EUS) are other techniques to be used side-by-side with ERCP. In the past decades, the development of magnetic resonance cholangiopancreatography (MRCP) has replaced the ERCP as a diagnostic method, and ERCP is used virtually only as a therapeutic procedure or as a tool to take targeted samples from biliary or pancreatic ducts (e.g. cytology, flow cytometry, bacterial samples, biopsies, and intraductal biopsies) or papilla biopsy.
In ERCP, the targeted duct, either the common bile duct (CBD) or the pancreatic duct (PD), is mainly cannulated with a guidewire and a sphincterotome (Testoni et al. 2016). Successful cannulation of the papilla Vateri is critical in ERCP. After insertion of the guidewire and the sphincterotome in the CBD or PD, such therapeutic procedures as stone removal, duct dilatations (benign or malign strictures), stenting, or ductal sampling can be performed.
The overall complication rates in ERCP are higher than in other endoscopic procedures, and post-ERCP pancreatitis (PEP) is the most common complication, with incidence rates of 3.5-9.7% (Dumonceau et al.
2020). PEP patients mainly have the mild form of the disease, but 11.4% of patients suffer from the severe form of PEP (Andriulli et al. 2007). Mortality in PEP is 0.7-3.6% (Kochar et al. 2015; Mutneja et al. 2021), and overall mortality in ERCP is 0.06-0.2% (Cotton et al. 2009; Kochar et al. 2015;
Mutneja et al. 2021). Several methods to decrease the PEP rate exist, the most important being the rectally administered non-steroidal anti- inflammatory drugs (NSAIDs) indomethacin and diclofenac, prophylactic pancreatic stenting, and aggressive hydration with lactated ringer`s solution (Dumonceau et al. 2020).
Difficult biliary cannulation, which varies between 1.4% and 49.5%
depending on the definition and patient selection, is one of the factors increasing the risk of PEP (Tse et al. 2017; Catalano, Linder, and Geenen 2004; Lee et al. 2015; Sugiyama et al. 2018; Zang, Zhang, and Gao 2014).
Repeated attempts to insert the guidewire into the CBD result in oedema and irritation to the papilla, leading to PEP. Different advanced cannulation
methods are required for biliary access in difficult cannulation. These methods include transpancreatic biliary sphincterotomy (TPBS), precut sphincterotomy, and double guidewire technique (DGW) (Testoni et al.
2016).
Primary sclerosing cholangitis (PSC) is a chronic liver disease affecting bile ducts and producing strictures and dilatations in the biliary tree.
Incidence of PSC is between 0.4/100 000 and 2.0/100 000 per year (Boonstra et al. 2013; Jepsen, Gronbaek, and Vilstrup 2015; Karlsen et al.
2017; Molodecky et al. 2011). PSC patients have almost a 400-fold risk of cholangiocarcinoma (CCA) compared with the general population (Boonstra et al. 2013). PSC patients in the HUS Endoscopy Unit undergo ERCPs for confirmation of diagnosis, for ductal sampling (dysplasia screening as a precursor for CCA), and as a therapy for biliary strictures. PSC progression to liver cirrhosis or repeated dysplasia findings in brush cytology indicate liver transplantation evaluation (Boyd, Vannas, et al. 2017). Furthermore, balloon dilatation and stenting provide alleviation of the symptoms of the disease. In HUS, approximately 500 ERCPs are performed annually for PSC patients.
PSC patients are at higher risk for PEP than other patients undergoing ERCP (Ismail et al. 2012; von Seth et al. 2015). The rectal administration of diclofenac 100 mg as an anti-inflammatory drug has been shown to reduce PEP risk and is recommended in the international guidelines. However, no clear data exist on whether PSC patients benefit from the prevention. PSC patients have also other autoimmune diseases such as inflammatory bowel disease (IBD) and autoimmune hepatitis (AIH). These patients use immunosuppressive thiopurines as a maintenance therapy. Thiopurines are risk factors for acute pancreatitis (AP) by unclear mechanism and may be risk factors also for PEP.
This thesis comprises four original studies of ERCP procedures targeting the CBD. The studies focus on complications, especially PEP risk, in patients undergoing ERCP. Two of the studies explore cannulation techniques and short- or long-term complications. The other two studies examine ERCPs performed on PSC patients. One study evaluates diclofenac as a PEP prevention and another study thiopurines as a possible risk factor for PEP.
ERCP was developed as a method to visualize the biliary and pancreatic ducts with X-rays. The first pancreatogram was published in 1968 (McCune, Shorb, and Moscovitz 1968), and the method rapidly developed as both a diagnostic and therapeutic procedure. Biliary sphincterotomy (BS) was described in 1974 as a method to remove biliary stones (Classen and Demling 1974; Kawai et al. 1974). Different diagnostic and therapeutic procedures have developed in the last 50 years. The diagnostic role of ERCP method was later challenged by the development of computer tomography (CT) and further by MRCP.
Newer interventions, such as EUS-guided cholangiography and EUS-guided biliary drainage, are methods to detect and treat biliary disorders. The direct ductal visualizing techniques of cholangioscopy and pancreatoscopy allow specific tissue sampling, biopsies, and lithotripsy (Tringali et al. 2015).
However, the ERCP remains an important therapeutic method for a targeted group of patients with biliary or pancreatic disorders.
Indications for ERCP according to the American Society for Gastrointestinal Endoscopy (ASGE), the European Society of Gastrointestinal Endoscopy (ESGE), and selected literature are listed in Table 1.
!#$-' /&+%#4+0/3(02#'&(20. #/& )5+&'-+/'3#/&
3'-'%4'&-+4'2#452'#$#,,'/'4#- 57$#5.'4#- #2-8'4#- #/'3 '4#- '2,/+.+42'4#-
$!-+!0/%
%0AB@C1B7D38/C<2713
7:32C1BAB=<3A
• 3FB@/1B7=<07:7/@G A>67<1B3@=B=;G=@:/@53 0/::==<27:/B/B7=<
• AB3<B7<54=@7<1=;>:3B3
@3;=D/:
• ;316/<71/::7B6=B@7>AG
• 16=:/<57=A1=>G/AA7AB32 :7B6=B@7>AG
7:7/@GAB@71BC@3A
• 0/::==<27:/B/B7=<
• AB3<B7<5
• @/27=4@3?C3<1G/0:/B7=<
>6=B=2G</;71B63@/>G7<
C<@3A31B/0:3BC;=C@A
&3@7>/>7::/@GBC;=C@
• >/>7::/07=>AG
• A>67<1B3@=B=;G/<2AB3<B7<5
• >/>7::31B=;G
&@7;/@GA1:3@=A7<5 16=:/<57B7A
• 16=:/<57B7A=0AB@C1B7=<
• AB@71BC@327:/B/B7=<AB3<B7<5
7:3:3/9=@AB@71BC@3
/4B3@AC@53@G
• 16=:31GAB31B=;G
• 07:7/@GB@/1BAC@53@G
• :7D3@@3A31B7=<
• :7D3@B@/<A>:/<B/B7=<
• 07:7/@GA>67<1B3@=B=;G
• AB@71BC@327:/B/B7=<
• AB3<B7<5
%#)*./%
(B@71BC@3A • 03<75< ;/:75</<B
• 16=:/<57=1/@17<=;/AB/57<5 0@CA61GB=:=5G4:=E
1GB=;3B@G ($(
• 16=:/<57=A1=>GD7AC/:7H7<5
AB@71BC@3AB/@53B32 A/;>:7<5 07=>AG
&@7;/@GA1:3@=A7<5 16=:/<57B7A
• 74#'&7A3?C7D=1/:
• AB/57<5
• B/@53B32A/;>:7<5 07=>AG
• 07:3A/;>:3A1GB=:=5G4:=E 1GB=;3B@G ($(
• 16=:/<57=A1=>GD7AC/:7H7<5 AB@71BC@3AB/@53B32 A/;>:7<5 07=>AG
$!-+!0/%
6@=<71>/<1@3/B7B7A )@3/B;3<B4=@AG;>B=;A
• >/7<
• >/<1@3/B712C1BAB@71BC@3A
• >/<1@3/B71AB=<3A
• >A3C2=1GAB2@/7</53
• >/<1@3/B=A1=>G:7B6=B@7>AG
%#)*./%
(B@71BC@3A
• 03<75< ;/:75</<B
• B/@53B32A/;>:7<507=>AG
• >/<1@3/B=A1=>GD7AC/:7H7<5 A/;>:7<5
<B@/2C1B/:>/>7::/@G
;C17<=CA<3=>:/A;
• >/<1@3/B718C7131GB=:=5G
• >/<1@3/B=A1=>GD7AC/:7H7<5 A/;>:7<5AB/57<5
CCA, cholangiocarcinoma; FISH, fluorescence in-situ hybridization; MRCP, magnetic resonance cholangiopancreatography; NGS, next-generation sequencing
ERCP is conducted with a flexible duodenoscope. Access to the desired duct, either the biliary or pancreatic duct, is achieved through the papilla major or papilla minor. A native papilla is one of the risk factors for PEP, and cannulation of the papilla might be challenging in ERCP. It is mainly performed with the guidewire, allowing the sphincterotome to enter the CBD or PD. After successful cannulation, different therapeutic procedures, such as stone extraction, dilatations of strictures, stenting, and tissue sampling, can be performed. The ESGE and the United European Gastroenterology (UEG)
recommend a quality ERCP to achieve a minimum 90% success in deep biliary cannulation (and 95% success in expert centres) of patients with normal anatomy and native papilla and to achieve a PEP rate lower than 10%
(target: lower than 5%) (Domagk et al. 2018). Use of a percutaneous transhepatic biliary drainage (PTBD) is advised if the ERCP for biliary obstruction, e.g. bile duct stones and acute cholangitis, fails (Manes et al.
2019). ERCP procedures can be graded to classes from 1 to 4 based on their complexity (Table 2) (Cotton et al. 2011; Schutz 2011).
!#$-' &+((+%5-48)2#&'#&014'&(20.0440/#/& %*5490440/'4#- %*549
- ! -*! 0-!
33>1/<<C:/B7=<=42C1B=47<B3@3AB;/7<>/>7::/=@A/;>:7<5 7:7/@GAB3<B@3;=D/:=@3F16/<53
7:7/@GAB=<33FB@/1B7=<J;;
)@3/B;3<B=407:7/@G:3/9A
)@3/B;3<B=43FB@/63>/B71AB@71BC@3A03<75<=@;/:75</<B
&:/13;3<B=4>@=>6G:/1B71AB3<BA 7:7/@GAB=<33FB@/1B7=<K;;
#7<=@>/>7::/1/<<C:/B7=<7<>/<1@3/A27D7AC;/<2B63@/>G '3;=D/:=47<B3@</::G;75@/B3207:7/@GAB3<BA
<B@/2C1B/:7;/57<507=>AG=@47<3<332:3/A>7@/B7=<
#/</53;3<B=4/1CB3=@@31C@@3<B>/<1@3/B7B7A
)@3/B;3<B=4>/<1@3/B71AB@71BC@3A
'3;=D/:=4>/<1@3/B71AB=<3AB6/B/@3;=07:3/<2J;;
)@3/B;3<B=467:/@BC;=C@A
)@3/B;3<B=403<75<07:7/@GAB@71BC@3A67:C;/<27<B@/63>/B71AB@71BC@3A
#/</53;3<B=4ACA>31B32A>67<1B3@=4%2272GA4C<1B7=<E7B6=@E7B6=CB
;/<=;3B@G
'3;=D/:=47<B3@</::G;75@/B32>/<1@3/B71AB3<BA
<B@/2C1B/:7;/535C7232B63@/>G35>6=B=2G</;71B63@/>G3:31B@=6G2@/C:71 :7B6=B@7>AG
'3;=D/:=4>/<1@3/B71AB=<3A7;>/1B32/<2 =@K;;
'3;=D/:=47<B@/63>/B71AB=<3A
&A3C2=1GAB2@/7</53=@<31@=A31B=;G
&/>7::31B=;G
'&/4B3@/,67>>:3>@=132C@3=@'=CF3<.0/@7/B@71AC@53@G
ERCP endoscopic retrograde cholangiopancreatography. One grade should be added (for a maximum grade of four) for procedures performed after normal working hours, after Billroth II gastrectomy, or for procedures that have previously failed.
Cannulation techniques have developed and changed in the past decades. A conventional biliary cannulation method was performed with a contrast- guided technique by injecting the contrast agent through a catheter or a sphincterotome to detect the location of the device in fluoroscopy (Freeman and Guda 2005). The technique disclosed whether the device was in the biliary tree or inside the PD (Cennamo et al. 2019). However, the method caused a considerable risk of PEP (Freeman et al. 2001). A wire-guided cannulation (WGC) technique was introduced in 1987 (Siegel and Pullano 1987). In the WGC technique, a sphincterotome with a guidewire is gently inserted into the papilla of Vater and the wire is guided through the CBD with fluoroscopy guidance (Baillie 2012; Freeman and Guda 2005). The WGC technique can be divided into “touch” and “no-touch” techniques, with the “touch” technique being more successful (Bassi et al. 2018). Conventional and sphincterotome techniques with guidewire were compared in a study in 2004, with a zero PEP rate noted with the sphincterotome technique (Lella et al. 2004). The results were similar in a systemic review; the WGC technique was found to reduce the complication rates, especially the PEP rate, and to increase the primary cannulation success rate (Tse, Yuan, and Moayyedi 2012). Thus, nowadays the WGC is the most preferred primary cannulation method in CBD cannulation (Freeman and Guda 2005; Testoni et al. 2016).
The Scandinavian Association for Digestive Endoscopy (SADE) research group published their criteria for difficult biliary cannulation in 2014 (Halttunen et al. 2014) after analysing 907 prospective biliary cannulations of native papillae. The definition of difficult cannulation was in WGC specified as 1) more than 5 min of attempting to cannulate, 2) more than five contacts with the papilla, or 3) more than one unintended PD passage. Since 2016, the ESGE has adopted this 5-5-2 definition (Testoni et al. 2016).
Another international consensus definition from Asia, Australia, and USA specifies difficult cannulation as 10 min of attempting to cannulate, more than 5 attempts, or more than one PD passage (Liao et al. 2017). However, according to previous studies, difficult biliary cannulation occurred in 1.4- 49.5% of cases, with an unstandardized definition (Tse et al. 2017; Catalano, Linder, and Geenen 2004; Lee et al. 2015; Sugiyama et al. 2018; Zang, Zhang, and Gao 2014).
In difficult CBD cannulation, repeated WGC attempts may increase PEP
in failure, different additional or advanced techniques can be performed as a rescue method. When the guidewire enters the PD, the DGW or TPBS technique can be performed. In addition, different rendezvous techniques (laparoscopic, PTBD, and EUS-guided) should be considered if the ERCP fails (Testoni et al. 2016).
Thus, the latest ESGE cannulation guidelines recommend early NK precut sphincterotomy, and if necessary, the DGW technique with prophylactic PD stent (Testoni et al. 2016). This recommendation concerning the DGW method with prophylactic PD stent is based on one prospective randomized study and two retrospective studies. All of the ERCPs were performed with contrast-guided cannulation (Ito et al. 2010; Ito et al. 2014; Nakahara et al.
2014). However, contrast-guided cannulation is a risk factor for PEP (Freeman et al. 2001). These studies cannot be generalized to everyday practice and the recommendation for prophylactic PD stents in DGW should be re-evaluated with high-quality studies. Furthermore, in the latest ESGE guidelines concerning PEP prophylaxis, the PD stents are recommended for selected patients at high risk for PEP. This aspect is evaluated in Section 2.2.1.1 under the heading PEP prophylaxis.
The TPBS method is the third option recommended in the algorithm provided by ESGE (Figure 3) (Testoni et al. 2016). However, one recent meta-analysis suggested that TPBS should be performed earlier in a difficult cannulation situation (Pécsi et al. 2017). Similar results were provided by a systematic review and a network meta-analysis of difficult biliary cannulations. The low-quality evidence supported the use of TPBS compared with the early precut technique, DGW, or cannulation over PD stent to increase biliary cannulation success. Early precut techniques and the TPBS led to lower PEP rates than the DGW method (Facciorusso et al. 2022).
In a prospective study evaluating the different macroscopic appearance of papillae, small and protruding or pendulous papillae were the most difficult to cannulate. However, no significant difference existed in PEP rates between the different manifestations of papillae (Haraldsson et al. 2019). Similar results were presented in two retrospective studies, where small or flat papillae and pendulous-type papillae were associated with longer cannulation or difficulties in cannulation success and higher PEP risk (Chen et al. 2020; Mohamed et al. 2021). When the papilla is situated inside the diverticulum, it is sometimes difficult to cannulate. It is possible to lift the diverticulum with biopsy forceps, to inject saline solution below the mucosa, or to use a pancreatic guidewire or stent to evert the papilla (Freeman and Guda 2005). The different papilla types can play a role in difficult biliary cannulation (Haraldsson et al. 2019).
It should be noted, however, that instead of repeated attempts with the same method, the endoscopist should be prepared to use different rescue methods selected individually. The 5-5-2 method was found feasible in a prospective study to help the endoscopist in decision-making when the primary cannulation technique failed (Ismail et al. 2019). However, the
safety of the procedure should be considered by weighing the risks against the benefits.
General cannulation techniques are presented in Table 3.
!#$-' +(('2'/4%#//5-#4+0/.'4*0&3(02%0..0/$+-'&5%4%#//5-#4+0/#&014'&
(20. )5+&'-+/'3!'340/+'4#-
-%(-3))0'/%*)/!$)%,0!. E7@35C7232 1=<B@/AB5C7232 -!0//%)#/!$)%,0!. <332:39<743>@31CB
<332:39<74347ABC:=B=;G A>67<1B3@=B=;3>@31CB 0% !2%-!%)+)-!/% 0/ 2=C0:35C723E7@3B316<7?C3
B@/<A>/<1@3/B7107:7/@GA>67<1B3@=B=;G
>/<1@3/B71AB3<B/AA7AB32B316<7?C3
/$!- >/>7::31B=;G
@3<23HD=CAB316<7?C3E7B6&)
*(5C72321/<<C:/B7=<
EUS endoscopic ultrasound; PTBD percutaneous transhepatic biliary drainage
#""#!
Precutting techniques include the conventional needle knife (NK) precut technique, sphincterotome precut and NK fistulotomy. The NK precut technique is defined as a stepwise incision of the duodenal mucosa starting at the upper margin of the papillary orifice (Freeman and Guda 2005). The incision opens the mucosa and the biliary sphincter. In the NK fistulotomy technique, the fistula is performed as a stepwise incision, starting directly over the roof of the papilla. The idea is to perform the cut either upward or downward to detect the underlying biliary sphincter. This technique avoids thermal injury to the PD orifice, thus decreasing the risk of PEP (Freeman and Guda 2005; Jang et al. 2020) Fistulotomy is sometimes helpful when primary cannulation fails and the distal CBD is wide. However, precut techniques should be performed in the early phase after WGC cannulation failure to diminish papilla manipulation (Maharshi and Sharma 2021).
##$ "#
Dumonceau et al. (1998) demonstrated the DGW technique 2.5 decades ago.
The idea of the technique is to introduce the second guidewire into the CBD if the first wire either accidentally or purposely enters the PD (Figure 1). The DGW method can be used, for example, in malformed CBD anatomy related to papilla morphology or neoplasia (Herreros de Tejada et al. 2009).
%#0-! =C0:35C723E7@3B316<7?C3)63/AA7AB/<BE7@37A7<B63&/<2B63 A>67<1B3@=B=;37A7<A3@B327<B=B63/4B3@B63A31=<2E7@375C@30G+
!=A93<A/:=
Several randomized and prospective studies have compared the DGW method with other methods, however, these studies have not shown the superiority of the DGW method. The PEP rate in a Spanish multicentre randomized study, comparing 97 patients with DGW with 91 patients with standard persistent cannulation technique was 17% vs. 8% and the success rate 47% vs. 56% (Herreros de Tejada et al. 2009). The DGW method was used in a Japanese study of 113 patients, yielding a success rate of 73%. In that study, inserting a pancreatic stent was a protective factor for PEP, with a PEP incidence of 4.7% compared with 22% in the no-stent group (Ito et al.
2008). A very low PEP rate (2.0%) without prophylactic pancreatic stents was achieved in a prospective study from Finland including 50 DGW patients. The cannulation success with the DGW method was, however, 66%, and the overall success rate after alternative techniques was 98% (Grönroos et al. 2011). A randomized study from Japan reported a higher success rate of 93% with DGW (n=27) than 58% with the conventional method (n=26) (Maeda et al. 2003). However, in a recent meta-analysis of seven randomized controlled trials, RCTs and 577 patients (289 patients in DGW and 288 patients in control groups), DGW was compared with other techniques (Tse et al. 2017). These techniques were persistent conventional cannulation (three studies) and other advanced techniques: precut with freehand fistulotomy, TPBS, PD stent, and DGW with or without PD stent. The authors found no significant difference in CBD cannulation success between DGW and the other randomized techniques (RR 1.04, 95% CI 0.87-1.24). DGW
significantly increased the PEP risk compared with other endoscopic techniques (RR 1.98, 95% CI 1.14-3.42). In addition, the DGW did not appear to improve cannulation success.
! " &!" "&
TPBS technique was demonstrated by Goff (Goff 1995). When the guidewire is in the PD, the sphincterotomy incision is made with a sphincterotome towards 11 or 12 o´clock to open the septum between the CBD and the PD (Figure 2, the green line in the smaller figure is the TPBS cut) (Halttunen et al. 2009). The cut reveals the lumen of the CBD. Eventually, the NK technique can be combined with TPBS, if necessary, towards 10 o`clock from the upper end of the TPBS cut (Halttunen et al. 2009) (Figure 2, the red line in the smaller figure is the additional NK cut). In a prospective study, the success rate with TPBS (n=116) was 85%. The overall complication rate in the TPBS group was 12% and the PEP rate 7.8%. There was no difference in the PEP rate when comparing biliary sphincterotomy and TPBS (Kahaleh et al.
2004). Only a few randomized studies comparing TPBS with other methods have been published. Two studies compared TPBS with NK precut or NK sphincterotomy, and success rates with TPBS vs. NK were 29/29, 100% vs.
24/34, 77% (p=0.01) and 70/73, 96% vs. 64/76, 84% (p=0.02) (Catalano, Linder, and Geenen 2004; Zang, Zhang, and Gao 2014).
%#0-! )@/<A>/<1@3/B7107:7/@GA>67<1B3@=B=;G)&(B316<7?C3)63/AA7AB/<BE7@37A 7<B63>/<1@3/B712C1B&/<2B63A>67<1B3@=B=;G7A>3@4=@;32B6@=C56B63 A3>BC;03BE33<B631=;;=<07:32C1B/<2B63&)635@33<:7<37AB63 )&(1CB/<2B63@32:7<3;/@9AB63/227B7=</:<332:39<743$!1CB75C@30G +!=A93<A/:=
A South Korean group published a randomized study of 71 patients comparing DGW (n=34) and TPBS (n=37). The CBD cannulation was defined as difficult after 10 minutes and 10 attempts. Within the limit of 10 extra attempts, the initial success rates and mean cannulation times in the DGW and TPBS groups were 79.4% vs. 78.4% (p=0.915) and 19 min vs. 20.5 min (p=0.732), respectively. The PEP rate was 38.2% and 10.8% in the DGW and TPBS groups, respectively (p=0.011) (Yoo et al. 2013). Another randomized study compared TPBS (n=34) with DGW (n=34), both with prophylactic pancreatic stenting. The PEP rates were similar, 2.9%, in both groups. TPBS succeeded most often (94.1% vs. 58.8%, p=0.001) (Sugiyama et al. 2018). A recent review article summarized that TPBS yields better cannulation success rates but has a similar complication profile as in the other methods (Pécsi et al. 2019).
" "!#" "
Other possible CBD cannulation techniques include papillectomy and cannulation over the pancreatic stent (Testoni et al. 2016). Depending on the cannulation situation, the anatomy of the ampulla, the presence of a
peripapillary diverticulum, and the expertise of the endoscopist, different advanced methods can be performed. If the conventional technique failed, the standard technique can be performed repeatedly, either by the same endoscopist in a different session or by another endoscopist. In addition, in a non-accessible situation, only a precut can be performed and the ERCP continued after two to three days. Rendezvous techniques include PTBD- assisted cannulation, where the catheter is inserted via the guidewire percutaneously to the biliary tract and further trans-papillarily to ensure drainage, and cannulation in the subsequent ERCP. In addition, EUS- assisted cannulation is another option in difficult ERCP. With EUS, it is possible to insert the guidewire trans-gastrically or trans-duodenally to the targeted duct antegrade and then ERCP is performed (Freeman and Guda 2005).
The algorithm of difficult biliary cannulation was published in 2016 in the ESGE guidelines (Testoni et al. 2016). This algorithm with small modifications is presented in Figure 3.
%#0-! :5=@7B6;=4274471C:B07:7/@G1/<<C:/B7=</2=>B324@=;B63(5C723:7<3A
;=274732)3AB=<73B/:'&3<2=A1=>71@3B@=5@/23
16=:/<57=>/<1@3B=5@/>6G*(3<2=A1=>71C:B@/A=C<2&)>3@1CB/<3=CA B@/<A63>/B7107:7/@G2@/7</53
'&! )&')$
%& $%" &$!&!
'$ "$$* '&! )&
'&%!$&&"&%!$
)&" $&'&"%%
'& '&!
!" $&
'& %$&!
" &$!&!*
"$'&
%&'!&!*
"$'&
"$'&
"&
! %$%%%&
$ +(!'%
'&!
($*%&
%%! %$
""&!*
& &! " $&')$ %$&!
! %$"$!"*&
" $&%& &
')$!'
& #'
!! '&' %&'!&!*
$ %" $&
%" &$!&!*$*
$&'&
%& &%%%&
"$!'$
ERCP is an invasive procedure. In a large register study, the overall perioperative complication rate was 2.5% (including bleeding and contrast material effusion) and the post-procedure complication rate 9.8% (including PEP and other complications) (Enochsson et al. 2010). Overall mortality was 0.06-0.2% (Cotton et al. 2009; Kochar et al. 2015; Mutneja et al. 2021).
ERCP complications are defined according to the consensus criteria (Table 4) (Cotton et al. 1991).
!#$-' 0/3'/353%2+4'2+#0(%0.1-+%#4+0/3#%%02&+/)400440/'4#-0440/
'4#-
%' * !-/! !1!-!
)-!/%/%. A3@C; >:/A;/
/;G:/A3/B:3/AB B7;3A*"$/B 6=C@A/4B3@B63 '&@3?C7@7<5 /2;7AA7=<=@
>@=:=<5/B7=<=4
>:/<<32/2;7AA7=<
B=2/GA
>/<1@3/B7B7A
@3?C7@7<5 6=A>7B/:7H/B7=<=4 2/GA
>/<1@3/B7B7A@3?C7@7<5 6=A>7B/:7H/B7=<4=@
2/GA=@6/3;=@@6/571
>/<1@3/B7B7A>6:35;=<
=@>A3C2=1GAB=@
7<B3@D3<B7=<
>3@1CB/<3=CA 2@/7</53=@AC@53@G/
<3324=@7<B3<A7D3 1/@3=@23/B6=4B63
>/B73<B
'!! %)# 1:7<71/:3D723<13=4 0:3327<573<=B 8CAB3<2=A1=>71 6/3;=5:=07<:3D3:
2@=>5<=<332 4=@B@/<A4CA7=<
B@/<A4CA7=<J C<7BA<=
/<57=5@/>671 7<B3@D3<B7=<=@
AC@53@G
B@/<A4CA7=<KC<7BA
=@7<B3@D3<B7=<
/<57=5@/>671=@
AC@53@G
!-"*-/%*) >=AA70:3=@=<:G D3@GA:756B:3/9=4 4:C72=@1=<B@/AB /53<BB@3/B/0:30G 4:C72A/<2AC1B7=<
4=@J2/GA
/<G2347<7B3
>3@4=@/B7=<
B@3/B32;3271/::G 4=@2/GA
;3271/:B@3/B;3<B4=@
;=@3B6/<2/GA=@
7<B3@D3<B7=<
>3@1CB/<3=CA=@
AC@571/:
)"!/%*) I43D3@/B 6=C@A
430@7:3=@A3>B71 7::<3AA@3?C7@7<5 2/GA=4 6=A>7B/:
A3>B71A6=19=@<332 4=@AC@53@G
B@3/B;3<B=@
3<2=A1=>71=@
>3@1CB/<3=CA 7<B3@D3<B7=<
ERCP endoscopic retrograde cholangiopancreatography; ULN upper limit of normal
However, the majority of ERCP complications are mild (Dumonceau et al.
2020), the most common complication being PEP. Other common complications include cholangitis, bleeding, and perforation. In addition, cardiopulmonary complications emerge in 4-16%; of these, 0.07-2.4% are clinically significant, resulting in mortality in 0.07%. Other miscellaneous complications (ileus, abscesses, other infections, pneumothorax) have also been reported (Chandrasekhara et al. 2017).
Incidence and mortality rates of common complications of ERCP are presented in Table 5.
!#$-' 2'-#4'&%0.1-+%#4+0/3#%%02&+/)40 5.0/%'#5'4#-
*(+'%/%*) )% !)! *-/'%/3
&/<1@3/B7B7A
6=:/<57B7A
:3327<5
&3@4=@/B7=<
PEP is the most common complication. Mostly—in 75% of cases—the PEP was diagnosed as mild (Cotton et al. 2009). However, the majority of patients diagnosed with PEP need hospital care, increasing the costs. A minority of patients require intensive care, which is extremely expensive. In addition, all complications cause personal suffering and PEP can increase the risk for further AP or PEP. Thus, the indication for ERCP must be carefully considered, weighing the risks and benefits for an individual patient. In addition, suitable methods to prevent PEP should also be considered and adapted to ERCP practice (Chandrasekhara et al. 2017; Dumonceau et al.
2020).
The consensus criteria of PEP diagnosis according to Cotton et al. (1991) are 1) a new or worsened abdominal pain with 2) elevated amylase or lipase value of more than three times the upper limit of normal (ULN) at 24 hours after the ERCP procedure and 3) a need for admission or prolongation of a planned admission at hospital. PEP can be categorized as mild, moderate, or
severe (Table 4) (Cotton et al. 1991). Other criteria for PEP diagnosis are the revised Atlanta criteria from 2012. The definition of PEP requires two of the following three symptoms/findings: 1) abdominal pain, typical for acute pancreatitis, 2) serum lipase or amylase levels at least three times the ULN, or 3) characteristic findings of AP on contrast-enhanced CT (Banks et al.
2013). The latest ESGE guidelines (Dumonceau et al. 2020) suggest determining the difficulty of PEP according to the revised Atlanta criteria (Banks et al. 2013). These criteria include the definition of organ failure and are divided into three levels: mild pancreatitis (no organ failure or no local or systemic complications exist), moderately severe pancreatitis (transient organ failure and/or local or systemic complication without persistent organ failure), and severe pancreatitis (persistent, single- or multiple-organ failure
>48 hours) (Banks et al. 2013).
According to the ESGE guidelines, the PEP rate is 3.5-9.7% (Dumonceau et al. 2020). The incidence rates are based on two large meta-analyses, where the PEP rates were 3.5% (16 855 patients, 21 studies, 95% CI 3.19-3.75%
(Andriulli et al. 2007)) and 9.7% (13 296 patients, 108 randomized trials, 95% CI 8.6-10.7% (Kochar et al. 2015)). Another large population-based register study presented a PEP rate of 2.7% (11 074 ERCPs) in unselected patients (Enochsson et al. 2010).
The detailed pathophysiological mechanism of PEP remains undetermined, and it is thought to be multifactorial (Tryliskyy and Bryce 2018). Repeated cannulation attempts can produce mechanical trauma to the papilla and to the pancreatic sphincter. Hydrostatic injury by pancreatic fluids and contrast media, and thermal injury by the electrical current used in sphincterotomy play a role as co-factors in the development of PEP (Freeman et al. 2001; Freeman and Guda 2004; Parekh et al. 2017). Difficult biliary cannulation with repeated cannulation attempts increases the risk of PEP to 10.8-16.2% (Dumonceau et al. 2014; Halttunen et al. 2014; Ismail et al. 2019; Lee et al. 2021; Testoni et al. 2010).
Commonly accepted patient- and procedure-related risk factors for PEP, according to the latest ESGE guidelines, are listed in Table 6. In different studies, the upper limit of of “younger age” varies from 40 to 65 years, however, in a large register-based study age of under 65 years was found to be a risk factor for PEP in univariate analysis (OR 1.68, 95% CI 1.45-1.94, p<0.001) (Syren et al. 2019).
!#$-' #4+'/4#/&120%'&52'2'-#4'&2+3,(#%4023(02&014'&(20.4*'
#/& )5+&'-+/'3*#/&2#3',*#2#'4#- 5.0/%'#5'4#-
/%!)/-!'/! !"%)%/!-%.&"/*- .-/%*
(CA>31B32(%
3;/:3A3F
&@3D7=CA>/<1@3/B7B7A
&@3D7=CA&&
&/B73<B@3:/B32:793:G@7A94/1B=@
.=C<53@/53
$=<27:/B323FB@/63>/B7107:32C1B
0A3<13=416@=<71>/<1@3/B7B7A
$=@;/:A3@C;07:7@C07<
<2AB/53@3</:27A3/A3
-*! 0-!-!'/! !"%)%/!-%.&"/*-
74471C:B1/<<C:/B7=<
&/<1@3/B715C723E7@3>/AA/53A
&/<1@3/B717<831B7=<
&@=132C@3@3:/B32:793:G@7A94/1B=@
&@31CBA>67<1B3@=B=;G
&/<1@3/B71A>67<1B3@=B=;G
7:7/@G0/::==<A>67<1B3@27:/B/B7=<
/7:C@37<07:7/@GAB=<33FB@/1B7=<
<B@/2C1B/:C:B@/A=C<2
SOD sphincter of Oddi dyskinesia; PEP post-endoscopic retrograde cholangiopancreatography pancreatitis
Experience of the endoscopist was reported as a significant factor in some studies (Cheng et al. 2006; Lee et al. 2020), but not in others (Freeman et al.
2001; Testoni et al. 2010; Vandervoort et al. 2002). A nationwide register data from the Netherlands showed that a higher lifetime number of ERCPs performed per endoscopist and a higher volume unit together were correlated with higher success rates. The success rate in native papilla ERCPs was 83.4% (4259/5106). OR for a failed procedure by high-volume endoscopists (>200 procedures per year) vs. low-volume endoscopists (<50 procedures per year) was 0.19 (95% CI 0.14-0.25, p<0.001) (Ekkelenkamp et al. 2015). In another study, high-volume endoscopists had 31% lower odds for complications than low-volume endoscopists (Keswani et al. 2017).
Repeated cannulation attempts are an independent risk factor for PEP (Ding, Zhang, and Wang 2015; Dumonceau et al. 2014; Freeman et al. 2001; Masci et al. 2003; Testoni et al. 2010; Wang et al. 2009). However, even if the cannulation is difficult (5-5-2 criteria) for an expert, only a low proportion (2.6%) of ERCP procedures fail (Halttunen et al. 2014).
&%!
The latest ESGE guidelines (Dumonceau et al. 2020) summarize the most important methods to decrease PEP risk: 1) NSAIDs prior to ERCP or 2) intensive hydration with lactated Ringer acetate, and if contraindications to previous methods, 3) a sublingual 5 mg single dose of glyceryl trinitrate. In addition, prophylactic PD stenting in certain situations and for patients at high risk for PEP is strongly recommended.
Non-steroidal anti-inflammatory drugs (NSAIDs)
Several randomized studies have investigated the role of NSAIDs as a PEP prophylaxis. In the pathophysiology of AP, the cascade of activation of digestive enzymes leads to autodigestive injury (Lankisch, Apte, and Banks 2015). Phospholipase A2, secreted from pancreatic acinar cells, is involved in the inflammatory process of AP (Mäkelä, Kuusi, and Schröder 1997). NSAIDs are potent inhibitors of phospholipase A and are downregulating pro- inflammatory agents (leukotrienes, prostaglandins, and platelet-activating agents) reducing inflammatory lesions and organ necrosis, thus affecting the development of AP (Lankisch, Apte, and Banks 2015; Mäkelä, Kuusi, and Schröder 1997; Serrano et al. 2019).
These different drugs include (either as a single drug or in combination with other agents) indomethacin (Andrade-Dávila et al. 2015; Döbrönte et al.
2014; Elmunzer et al. 2012; Levenick et al. 2016; Patai et al. 2017;
Sotoudehmanesh et al. 2007), diclofenac (Cheon et al. 2007; Hauser et al.
2017; Katsinelos et al. 2012; Khoshbaten et al. 2008; Lua, Muthukaruppan, and Menon 2015; Murray et al. 2003; Otsuka et al. 2012; Park et al. 2015), and naproxen (Mansour-Ghanaei et al. 2020) administered rectally, intravenously, orally, or intramuscularly. In rectal administration, compared with the oral route, the bioavailability of the drug is faster because of more rapid absorption and bypassing of the gastro-hepatic axis (Lyu et al. 2018). It has been shown that the other forms (intravenous, intramuscular, or peroral) of drug dosage do not influence PEP risk (Lyu et al. 2018; Rustagi and Njei 2015; Serrano et al. 2019). Contraindications to rectal administration of NSAIDs include renal failure, allergy or pseudoallergy, aspirin-intolerant asthma, and pregnancy after 30 weeks of gestation (Dumonceau et al. 2020).
Several meta-analyses have concluded that rectally administered indomethacin or diclofenac reduce the risk of PEP, especially in high-risk patients (Dai, Wang, and Zhao 2009; Ding et al. 2012; Serrano et al. 2019;
Sethi et al. 2014; Sun et al. 2014; Yaghoobi et al. 2013; Yang et al. 2017;
Yuhara et al. 2014). The number needed to treat (NNT) was 11-20 in overall risk reduction and 27-34 in moderate and/or severe PEP risk reduction (Ding et al. 2012; Patai et al. 2017; Sethi et al. 2014; Yaghoobi et al. 2013). Absolute risk reduction was 5.9% (Ding et al. 2012). By contrast, no reduction in PEP rates was found in one recent meta-analysis that included only low- or average-risk patients (Feng et al. 2017). Rectal NSAIDs, however, are
Prophylactic pancreatic stents (PPS)
The ESGE and ASGE guidelines both recommend prophylactic pancreatic stenting in selected high-risk patients for PEP prevention (Chandrasekhara et al. 2017; Dumonceau et al. 2020). The benefit of prophylactic stenting has been shown in several meta-analyses. (Choudhary et al. 2011; Mazaki et al.
2014; Shi et al. 2014; Vadala di Prampero et al. 2016). The number needed to treat in order to prevent one PEP was eight (Choudhary et al. 2011)
Short (3-5cm) 5-Fr stents are the most recommended (Afghani et al.
2014; Dumonceau et al. 2020). For prophylactic pancreatic stenting, ESGE suggests the use of a short 5-Fr stent without an internal flange (Dumonceau et al. 2020). Interestingly, a Swedish nationwide, register-based study concerning PPS revealed that larger than 5-Fr stents, compared with smaller diameter stents, reduce PEP risk (Olsson et al. 2017). In addition, stents over 5-Fr and longer than 5 cm were the most effective in PEP prevention.
However, no detailed data concerning the indication for stenting or the stent type were provided (Olsson et al. 2017). A flange or a pigtail should be present on the duodenal side of the stent for easier removal and to prevent migration inside the duct. Passage of the stent from the PD should be evaluated within 5 to 10 days after placement (with x-rays), and retained stents should be removed endoscopically. Stents should remain in the PD for at least 12-24 hours (Dumonceau et al. 2020). In addition, biodegradable stents offer the advantages of requiring no follow-up imaging or ERCPs for stent removal (Anderloni et al. 2020).
No clear data indicate that after proper TPBS a PPS is beneficial (Dumonceau et al. 2014), and further studies are needed to address this issue. One RCT concerning DGW patients with PPS was performed with different ERCP standards (as contrast-guided cannulation without NSAID prophylaxis), and thus, the results in PEP prevention with PPS are not fully generalizable (Ito et al. 2010). Only 2.0% PEP rates were presented in a prospective study concerning Finnish DGW patients without PPS (Grönroos et al. 2011). A recent meta-analysis comparing the different methods to prevent PEP concluded that PPS are the most efficient method (Njei et al.
2020). In a cost-effectiveness analysis, the prophylactic stenting seemed to be beneficial in high-risk patients (suspected SOD, history of PEP or recurrent AP, repeated PD opacification during the ERCP, and in cases of precut sphincterotomy, pancreatic sphincterotomy, or balloon dilatation of the biliary sphincter) (Das et al. 2007). Furthermore, a recent cost- effectiveness analysis suggested using PPS in high PEP-risk patients and in patients with contraindications to rectal indomethacin or aggressive hydration (Thiruvengadam et al. 2022).
PPS should be considered in high-risk patients with repeated cannulation attempts in difficult cannulation when the guidewire is in the PD (Elmunzer 2017). In addition, PPS are strongly recommended in endoscopic papillectomy according to the ESGE guidelines (Vanbiervliet et al. 2021).
However, a significant risk for both complications (stent migration, PD
perforation) and PEP, as high as 35% in attempted but unsuccessful stent placement, has been published (Choksi et al. 2015; Elmunzer 2017). Thus, evaluation of the need for PPS must be performed individually.
Other factors
Recent ESGE guidelines provide an algorithm for PEP prophylaxis (Dumonceau et al. 2020). Intravenous Ringer`s lactate solution is an alternative to rectal NSAIDs when contraindications to these drugs exist.
This high-volume hydration should be administered at 3 mL/kg/hour.
However, in elderly patients it should be noted that, for example, congestive heart failure is a contraindication for massive hydration. If a pancreatic stent is applied, there is no need for further hydration. If no pancreatic stent is applied, hydration is recommended to continue for 8 hours with one extra bolus, 20 mL/kg, just after the ERCP (Dumonceau et al. 2020). One meta- analysis with three randomized studies concluded that aggressive hydration can prevent PEP (overall, and development of moderate or severe PEP) and shorten hospitalization. However, no congestive heart failure patients, chronic kidney disease patients, or elderly patients were included in these studies (Wu et al. 2017). In another meta-analysis with 7 randomized studies, in 620 patients in the aggressive hydration group vs. 607 patients in the control group, the OR for PEP was 0.41 (95% CI 0.27-0.62), showing that aggressive hydration is beneficial when NSAIDs are contraindicated (Wang et al. 2021). In a recent randomized multicentre study, rectal NSAID and aggressive hydration (20 mL/kg/first hour and 3 mL/kg for the next 8 hours) together vs. rectal NSAID alone did not affect PEP rates (30/388, 7.7% vs.
39/425, 9.2%, p=0.53) (Sperna Weiland et al. 2021). Aggressive hydration for 24 hours increases the costs of ERCP procedures; most patients are treated as out-patients, but for hydration, patients need a place in the ward.
However, in cases of high PEP risk if PPS or rectal NSAIDs are unavailable or contraindicated, aggressive hydration was found to be preferable and cost- effective in prevention (Thiruvengadam et al. 2022).
Another possible method to prevent PEP is a single dose of 5 mg of glyceryl nitrate when high-volume hydration or NSAIDs are contraindicated.
In a meta-analysis, sublingual glyceryl nitrate was found to be beneficial for PEP prevention. Overall incidence of PEP was reduced (RR 0.67; 95% CI 0.52-0.87), however, no effect on moderate to severe PEP was observed (Ding et al. 2013). Furthermore, a large meta-analysis compared different drug therapy with placebo in PEP prevention (Lyu et al. 2019). Gabexate, glyceryl nitrate, somatostatin, ulinastatin, and NSAIDs (indomethacin and diclofenac) were more effective than placebo, and other drugs (ocreotide, allopurinol, nafamostat) were similar to placebo. In ocreotide/somatostatin, the effect is probably dose-dependent. However, these other drugs are not recommended in the ESGE guidelines (Dumonceau et al. 2020).
Other possible factors, e.g. size of the guidewire (0.025-inc or 0.035-inc, hydrophilic guidewires) in native papilla cannulation, were not significant in
the risk of PEP (Halttunen and Kylanpaa 2013). However, data on PEP prevention are lacking.
Cholangitis is determined as a new onset temperature >38°C for more than 24 hours, combined with cholestasis (Cotton et al. 2010). Risk of post-ERCP cholangitis (PEC) is increased in incomplete biliary drainage, in liver hilar obstruction, and in PSC or cholangioscopy (Dumonceau et al. 2020). In acute cholangitis, antibiotic treatment and ERCP with biliary drainage are recommended (Khashab et al. 2015). Acute cholangitis is classified according to the Tokio guidelines as mild (grade I, responds to initial medical treatment), moderate (grade II, no response to initial medical treatment, but no organ dysfunction), and severe (grade III, with acute organ dysfunction) (Kiriyama et al. 2018). In incomplete biliary stone extraction, bile drainage with plastic stents is recommended (Manes et al. 2019).
Prevention of cholangitis
Routine antibiotics are not recommended in ERCP. In a meta-analysis of randomized trials of altogether 1389 patients, no difference existed in PEC rates between the antibiotic prophylaxis group and the controls (Bai et al.
2009). A systematic review concluded that antibiotics prevent bacteraemia and might prevent cholangitis and septicaemia in patients without previous cholecystitis, cholangitis, or severe acute pancreatitis (Brand, Bizos, and O'Farrell 2010). However, ESGE and ASGE recommend prophylactic antibiotics in biliary obstruction (absence of clinical cholangitis) and suspicion of incomplete drainage to prevent PEC. In addition, temporary biliary stents are recommended to provide bile flow in situations of incomplete stone removal (Buxbaum et al. 2019; Khashab et al. 2015;
Dumonceau et al. 2020; Manes et al. 2019).
Post-ERCP bleeding is defined as haematemesis and/or melena, or haemoglobin decrease of more than 20 g/L (Cotton et al. 2010). Bleeding is mainly a consequence of endoscopic sphincterotomy. The risk factors associated with post-ERCP bleeding include anticoagulant intake, low platelet count (< 50 E9/L), liver cirrhosis, dialysis for end-stage renal disease, intraprocedural bleeding, unsuccessful cannulation with precut sphincterotomy, and low endoscopist experience (Dumonceau et al. 2020).
Anti-platelet agents together with different ERCP interventions were evaluated in a prospectively collected database study; only a pull-type sphincterotomy was associated with a higher risk of bleeding (Oh et al.
2018). Fully covered metal stents have been used as a treatment in post- sphincterotomy bleeding, either as a primary control or after conventional
