Technical overview of difficult biliary cannulation in endoscopic retrograde cholangiopancreatography: A guide for the endoscopist

Abstract

Endoscopic retrograde cholangiopancreatography (ERCP) is one of the most technically challenging procedures in the field of therapeutic endoscopy and remains the standard of care for the management of extrahepatic biliary obstruction (both benign and malignant causes). The prerequisite for defining the success of this procedure is to achieve deep biliary cannulation and the guidewire assisted technique is the most commonly performed procedure. A latest meta-analysis of 15 randomized controlled trials in 4426 patients reported the unweighted pooled primary cannulation success rate (using this technique) of 85.4% (at the hands of experienced endoscopists), with a failure of approximately 5%-20% overall, and in less than 5% at high volume centers. As a result, achieving deep cannulation is a substantial barrier to success of ERCP, both for experts and novices alike. Despite improvements in endoscopic technologies and accessories, development of advanced endoscopy fellowship programs, and advances in ancillary imaging techniques, biliary cannulation in ERCP can still be unsuccessful in up to 20% of patients, even in referral centers. Once cannulation has been deemed to be difficult, the risk of post-ERCP pancreatitis and technical failure inherently increases. A number of factors, including endoscopist experience and patient anatomy, have been associated with difficult biliary cannulation (DBC), but predicting a case of difficult cannulation a priori is often not possible. Numerous techniques such as pancreatic guidewire and stenting, early pre-cut, and endoscopic ultrasound-guided rendezvous may be employed when standard approaches fail. Data regarding the rate of success and adverse events of these techniques have been variable, though most studies suggest that pancreatic duct stenting generally reduces the rate of post-ERCP pancreatitis in instances of DBC. Here we provide a technical review on DBC (with images) and discuss how the choice of which techniques to employ and how to best employ them should be individualized and take into account the skill of the endoscopist, the disorder being treated, the anatomy of the patient, and the available biomedical literature.

Key Words: Endoscopic retrograde cholangiopancreatography; Bile duct; Needle knife; Double guidewire; Rendezvous; Transpancreatic sphincterotomy; Pancreatitis

Core Tip: Difficult biliary cannulation remains a significant challenge during endoscopic retrograde cholangiopancreatography, even in expert hands and high-volume centers, which has been shown to directly impact technical success and patient outcomes. This comprehensive review highlights current strategies available in hand for managing such clinical scenarios when encountered, including early identification of risk factors, optimized guidewire techniques, and the judicious use of advanced approaches such as pre-cut sphincterotomy, double guidewire technique, transpancreatic septotomy and endoscopic ultrasound-guided rendezvous technique. Understanding the timing and escalation to alternative methods or referral is critical. A structured, stepwise approach by practicing endoscopists enhances technical success and reduces complications in such challenging clinical scenarios.

INTRODUCTION

Endoscopic retrograde cholangiopancreatography (ERCP) is a complex therapeutic endoscopic procedure that plays a critical role as a first line modality for the management of pancreaticobiliary disorders. ERCP has evolved from a primarily diagnostic to a predominantly therapeutic tool. The rate limiting step of the procedure in ERCP is selective biliary cannulation, through which common bile duct (CBD) access is achieved for subsequent interventions. ERCP still carries a failure rate of 5%-20% even in high volume centers and expert endoscopists alike. Difficult biliary cannulation (DBC) also increases the risk of complications like post ERCP pancreatitis (PEP), sepsis/infection, perforation, and bleeding. The European Society of Gastrointestinal Endoscopy guidelines have defined DBC in an intact papilla as any procedure in which the duration of cannulation attempt exceeded 5 minutes or 5 attempts, or a procedure with more than one unintentional pancreatic duct (PD) cannulation or opacification[1]. The incidence of DBC can vary based on gastrointestinal anatomy, underlying pathology (benign or malignant), hospital case volume and operator expertise. Factors contributing to DBC include anatomical variations (small papilla, intra-diverticular papilla, protruding papilla), inflammatory conditions (chronic pancreatitis, prior surgery), and papillary distortion due to tumors or prior interventions. The identification and management of these factors is important in achieving successful technical and clinical outcomes.

Numerous advanced cannulation strategies have been developed over time. These include pre-cut sphincterotomy [needle-knife papillotomy (NKPP) or needle knife fistulotomy (NKF), double-guidewire (DGW) technique, and trans-pancreatic sphincterotomy (TPS)]. Recently developed techniques include endoscopic ultrasound (EUS)-guided rendezvous (RV)[2] technique and other hybrid RV techniques. The choice of technique is influenced by the specific clinical context, papillary anatomy and local expertise. The timing of transitioning from standard to advanced techniques is important. Early implementation of rescue techniques potentially improves outcomes and minimizes complications, especially PEP. Prophylactic measures against PEP, such as PD stenting and rectal non-steroidal anti-inflammatory drugs, are now standard practice in high-risk cases. Balancing the need for successful cannulation with the imperative to minimize harm requires a nuanced understanding of the available techniques, their indications, limitations, and evidence base.

This review aims to provide a comprehensive overview of DBC during ERCP and current best practices for management, with special emphasis on evidence-based techniques (with images). By consolidating current knowledge and recent advances, this article seeks to guide endoscopists in optimizing outcomes for patients undergoing ERCP in the setting of DBC. We will first describe the rescue techniques available for cases with DBC and normal anatomy. Thereafter, we will discuss the options available for management of cases with surgically altered anatomy.

ERCP IN NORMAL GASTROINTESTINAL ANATOMY

Rescue techniques used for DBC in cases with normal anatomy can be broadly categorized into three categories. Free hand technique involves directly cutting over the papilla and intraduodenal portion of the CBD, without the stability of a guidewire. Over pancreatic guidewire implies cannulation with the assistance of a guidewire placed in the PD. Needle knife (NK) papillotomy over a PD stent can be placed in either category. Although, truly a free hand technique, presence of the PD stent makes this technique more stable and safer than other free hand techniques. The third category is EUS guided RV. The feasibility of this procedure does not depend on papillary anatomy, but requires considerable expertise for safe and successful execution.

FREE HAND TECHNIQUES

Pre-cut papillotomy/fistulotomy

The term pre-cut is used when an incision is directly made either on the ampulla of Vater or on the intra duodenal segment of the CBD, in order to gain access into the CBD. NK are catheters where the tip is mounted with a 5 mm cutting wire. Length of the wire outside the distal tip of the catheter can be manually controlled. Electrosurgical current is passed through the wire and the catheter is manipulated so as to incise the papillary mucosa. The incision on the papillary mound can either be made in “down to up” or “up to down” direction. In “down to up” technique, the cutting wire is extended for 2-3 mm beyond the catheter tip and small incremental incisions are made in the 11-12 o’clock direction. The movement of the catheter tip is controlled using shaft of the scope, the up knob and the elevator. It is recommended to not extend the incision beyond the lower two third of the papillary mound, as the risk of bleeding and perforation increase. The plane beyond the incision can be explored using either the blunt tip of the catheter, carbon dioxide insufflation or water irrigation.

The biliary orifice appears like a pale onion skin like tissue, from which bile flow may be noted. The biliary orifice may be present at any position within the incision but is usually not present at its superior tip. The incision’s depth and length can be increased if the orifice is not identified after the initial cut. However, caution must be applied as deeper incision can result in inadvertent retro-duodenal perforations. Once the biliary orifice is identified, it can be incised to expose the reddish biliary mucosa[2,3].

Another method of pre-cut papillotomy involves using the Erlangen type pre-cut papillotome[4]. This papillotome has a short monofilament cutting wire at its distal end and a leading tip of 1 mm. Risk of injury to the PD may be lower due to the Teflon covered leading tip, and the cut is also directed cranially. Prospective study by Binmoeller et al[4] demonstrated a 91% successful cannulation rate at the initial attempt and a PEP rate of 2.7% with this papillotome.

In the “up to down” technique, the incision is started near the roof of the papillary mound and extends toward the papillary opening in the 5 o’ clock direction. Since the upper limit of the incision is predetermined in this technique, the risk of perforation is less. A modification of NKs has been described where submucosal injection of 6% heta-starch and methylene blue was used to create more space and help in delineation of various anatomical landmarks[5].

Successful cannulation rate with NKPP technique ranges from 70%-93% in various randomized controlled trials (RCTs)[6-10]. Most common complications are PEP and bleeding. Table 1 describes RCTs which have compared NKPP with other techniques including TPS, NKF and repeated cannulation over guidewire. The RCTs were quite heterogenous in term of the definition used for DBC and the etiology of biliary obstruction. However, acceptable cannulation rates with NK techniques further highlight their effectiveness in different scenarios. Furthermore, in expert hands, the complication rates are less than 10% and similar to other techniques. A meta-analysis of five prospective and eight retrospective studies compared the success and complication rates of NK papillotomy and trans pancreatic sphincterotomy[11]. NKPP had significantly lower success rates and higher bleeding rates than TPS, but there was no significant difference in the risk of PEP, perforation or overall complication rates. Maharshi and Sharma[12] conducted an RCT to compare the incidence of PEP and successful cannulation between primary NKPP and early NKPP. While cannulation rates were similar in both the groups, incidence of PEP was significantly lower in the primary NKPP group.

Table 1 Studies (randomized controlled trials) comparing needle knife vs other techniques for difficult biliary cannulation during endoscopic retrograde cholangiopancreatography, n (%)/mean ± SD.

Ref.	Definition of DBC	Endoscopist experience	Sample size	Baseline characterstics; age	Baseline characterstics; gender (male %)	Groups (n)	Cannulation rate	Bleeding	Perforation	PEP	Comments
Mavrogiannis et al[6], 1999	Failure to cannulate for more than 15 minutes by fellow and more than 20 minutes by experienced endoscopist; PD cannulation up to 5 times	Not defined	153	NKPP: 62.1 (29-83); NKF: 59.3 (20-91)	NKPP: 24 (30); NKF: 22 (29)	NKPP: 79; NKF: 74	NKPP: 58 (73.1) vs NKF: 56 (75.67)	NKPP: 4 (5.06); NK: 5 (6.755)	NKPP: 2 (2.530); NKF: 2 (2.27)	NKPP: 6 (7.59); NKF: 0	1 death in NKPP-severe PEP
Catalano et al[7], 2004	Not defined	Not defined	63	N/A	N/A	TPS: 3; NKPP: 33	TPS: 29 (94) vs NKPP: 24 (75)	-	-	-	21 (61) had PD stents in NK group
Zhang et al[8], 2022			188; prospective cohort study based on papilla type			NKPP: 75; NKF: 113	NKPP: 68 (90.7) vs NK: 104 (92)	NKPP: 1 (1.3); NKF: 3 (2.6)	NKPP: 0; NKFF: 1 (1.3)	NKPP: 5 (6.65); NKF: 6 (5.3)	-
Tang et al[9], 2005	Failure to cannulate within 7 minutes by an endoscopy fellow or within 5 minutes by experienced endoscopist	Not defined	62	NKPP: 64.6 ± 13.3; RSG: 67.2 ± 12.7	NKPP: 15 (46.9); RSG: 14 (46.7)	NKPP: 32; RSG: 30	NKPP: 24 (75) vs RSG: 22 (73.3)	NKPP: 3 (12.5); RSG: 0	NKPP: 0; RSG: 0	NKPP: 0; RSG: 1 (4.5)	-
Zhou et al[10], 2006	Failed cannulation after 10 minutes or pancreatic duct entry > 3 times		91			NKPP: 43; RSG: 48	NKPP: 39 (90.7) vs RSG: 36 (75)	-	-	-	Complication rate; NKPP: 9.3; RSG: 14.6
Maharshi and Sharma[12], 2021	> 2 failed cannulation attempts	Experience > 25 years; > 1000 ERCP per year	303	NKPP: 47.5 ± 14.5; primary NKPP: 48.2 ± 15.4	NKPP: 126 (41.5); primary NKPP: 61 (40.1)	NKPP: 152; primary NKPP: 151	NKPP: 141 (92.7) vs primary NKPP: 140 (92.7)	NKPP: 3 (1.95); primary NKPP: 4 (2.6)	NKPP: 1 (0.66); primary NKPP: 1 (0.67)	NKPP: 8 (5.2); primary NKPP: 1 (0.67)

DBC: Difficult biliary cannulation; PD: Pancreatic duct; PEP: Post endoscopic retrograde cholangiopancreatography pancreatitis; NKPP: Needle knife precut papillotomy; NKF: Needle knife fistulotomy; NK: Needle knife; N/A: Not applicable; TPS: Trans-pancreatic sphincterotomy; RSG: Repeated single guidewire cannulation; ERCP: Endoscopic retrograde cholangiopancreatography.

NKF

The procedure of NKF is a free-hand technique in which an incision is made on the intra duodenal of the CBD at the site of maximal protuberance. The incision leads to formation of a choledochoduodenal fistula, away from the papilla. The advantage of this technique is that there is no manipulation at the papilla, hence decreasing the risk of PEP. Access to the CBD from a choledochoduodenal fistula was first described by Osnes et al[13]. They had described two cases of choledocholithiasis with spontaneous choledochoduodenal fistula. The fistulous site was extended using a diathermy snare, cholangiogram was obtained and clearance was done from the same site.

NKF can be safely performed on papillae with a prominent intra-duodenal segment, and should be avoided in cases with small papilla (diameter < 3mm; type 2). The blunt tip of the catheter should first be used to palpate the intraduodenal segment of the bile duct. The bile duct feels like a mobile cord which is firmer than the surrounding submucosal tissue[14]. Incision can be made in either a down-up or up-down direction. However, the incision should not be extended too superior up to the edge of the duodenal bulge, as the risk of perforation increases. Depth of the incision should be controlled by varying the length of the diathermy wire. After the initial incision, blunt dissection can be done using the catheter tip, air or water. The biliary sphincter has an onion skin appearance and may or may not be visible. On further incision, biliary mucosa gets exposed and it has a red velvety appearance; flow of bile may be noted[15]. If biliary structures are not clearly visible, guidewire cannulation can be attempted from the papillary orifice. Protrusion of the wire from the fistulotomy site indicates adequate depth of incision[16].

Artifon et al[17] described the use of an 18-G needle to puncture the intra duodenal part of the CBD. They achieved biliary cannulation in 25 of 28 (89%) patients, complications included two cases of perforation and one with minor bleed. A modification of the NKF procedure has been described where an endoscopic sub-mucosal dissection knife is used to make the puncture[18,19].

Although NKF is reserved for cases with DBC, there have been studies which advocate for upfront NKF in patients with favorable anatomy. In a prospective study of patients at high risk for PEP (age < 60 years, female, CBD diameter < 9 mm), selective CBD cannulation rate with the primary NKF technique was 96.3% (53/55). None of the patients developed pancreatitis, while there were 2 cases (3.6% of minor bleed) and one case of minor perforation[20]. A meta-analysis of three RCTs and one prospective trial showed that selective cannulation rates with primary NKF was similar to that of conventional wire guided techniques, however, odds of PEP were significantly lower[21]. Another meta-analysis which compared the outcomes of primary NKF with rescue sphincterotomy techniques also found a significantly lower incidence of PEP in the primary NKF group, with similar rates of successful CBD cannulation and other complications between the two groups[22]. A meta-analysis comparing NKPP and NKF found similar cannulation rates between the two techniques (86.9% vs 91.7%). While the overall complication rate was lower with NKF, rate of PEP, bleeding and perforation was similar between both the groups[23].

OVER PANCREATIC GUIDEWIRE TECHNIQUES

Double guidewire technique

This technique was first described by Dumonceau et al[24]. In the event on an inadvertent PD cannulation during ERCP, the guidewire is left within the PD. The guidewire serves the purpose of stabilizing the papilla, occupying the PD opening and straightening the intra duodenal portion of the CBD. After leaving the guidewire in the PD, biliary cannulation is re attempted using a sphincterotome and a second guidewire. The endoscopist should aim to direct the wire toward the 11-12 o’clock direction for successful biliary cannulation. Once biliary cannulation is achieved, the guidewire from the PD can simply be removed. However, we recommend placing a temporary single pigtail 5 Fr plastic stent, so as to decrease the incidence of PEP (Figure 1).

Figure 1 Double guidewire technique for difficult biliary cannulation. A and B: Inadvertent pancreatic duct (PD) cannulation during conventional endoscopic retrograde cholangiopancreatography (guidewire entering PD) (A) on endoscopic image and on fluoroscopic image (B); C and D: With the guidewire in place in PD, sphincterotome used to selectively cannulate the bile duct (double guidewire technique) (C), this was which was confirmed on fluoroscopic image (D); E and F: Once bile duct cannulation is confirmed, deep cannulation was achieved and was successful (E), as confirmed on fluoroscopic image (F).

Table 2 describes the findings of RCTs that compared DGW technique with other methods including TPS, and repeat single guidewire cannulation[25-30]. In various RCTs the rate of successful cannulation using DGW ranges from 47%-92%[29,31]. The threshold for DBC varied significantly between studies. Insertion of a pancreatic stent was not uniformly followed in all studies, and the true risk of PEP after placement of a PD stent is not yet clear. A meta-analysis of seven RCTS studied the risk of PEP with DGW technique as compared to conventional cannulation and other rescue techniques. Risk of PEP was similar with DGW and conventional cannulation method, however, the risk of PEP as significantly higher with DGW as compared to pre-cut sphincterotomy[31-34]. Overall rate of successful CBD cannulation and risk of bleeding were similar between DGW and other groups.

Table 2 Studies (randomized controlled trials) comparing over pancreatic guidewire techniques for difficult biliary cannulation during endoscopic retrograde cholangiopancreatography, n (%)/mean ± SD.

Ref.	Definition of DBC	Endoscopist experience	Sample size	Baseline characterstics; age, years	Baseline characterstics; gender (male %)	Groups (n)	Successful CBD cannulation	Bleeding	Perforation	PEP	Comments
Eminler et al[25], 2019	Maximum of 5 guidewire passages into PD or cannulation attempt duration 5 minutes	Experts performing > 500 cases/year	100	WGC-PS: 52.9 ± 19.8; DGW: 56 ± 18.08	WGC-PS: 16 (32); DGW: 24 (48)	WGC-PS: 50; DGW: 50	WGC-PS: 27 (54); DGW: 45 (90)	WGC-PS: 0; DGW: 1 (2)	WGC-PS: 1 (2); DGW: 0	WGC-PS: 1 (2); DGW: 1 (2)	PD stent placed
Laquière et al[26], 2022	Guidewire passage into PD-1	Experts performing > 300 cases/year	142	DGW: 66.9 ± 15.6; RSG: 67.3 ± 15.8	DGW: 35 (51); RSG: 37 (50)	DGW: 68; RSG: 74	DGW: 57 (84); RSG: 37 (50)	DGW: 3 (4); RSG: 0	DGW: 1 (1); RSG: 0	DGW: 1 (1); RSG: 4 (5);	Early DGW, PD stent placed
Angsuwatcharakon et al[27], 2012	Inability to cannulate within 10 minutes by expert endoscopist	Expert performing > 300 cases/year for > 10 years	44	DGW: 66 ± 13.2; NKF: 64 ± 13.9	DGW: 13 (57); NKF: 10 (48)	DGW: 23; NKF: 21	DGW: 17 (73.9); NKF: 17 (81)	DGW: 2; NKF: 2	DGW: 0; PS: 0	DGW: 4; PS: 2	No PD stent; no PD stent
Sasahira et al[28], 2015	Cannulation attempt 10 times, up to 10 minutes	29 experts; 16 trainees	280	N/A	DGW: 74 (71.8); RSG: 67 (69)	DGW: 103 (75); RSG: 96 (70)				DGW: 27 (20); RSG: 23 (17)	PD stent: 18 DGW, 16 RSG; PD stent placed DGW: 25 (18); RSG: 22 (16)
Herreros de Tejada et al[29], 2009	No cannulation after 5 attempts	Experience from 6 years to 30 years	188	DGW: 69.5 ± 15.6; RSG: 65.8 ± 15.5	DGW: 38 (39); RSG: 38 (42)	DGW: 97; RSG: 91	DGW: 46 (47); RSG: 51 (56)	DGW: 0; RSG: 5 (6)	DGW: 1; RSG: 1	DGW: 13 (17); RSG: 7 (8)
Kylänpää et al[30], 2021	No cannulation after 5 minutes, > 5 contacts with papilla or > 1 unintended entry into PD	Not defined	203	DGW: 68 ± 24.92; TPS: 66 ± 21.97	DGW: 40 (40.4); TPS: 58 (55.7)	DGW: 99; TPS: 104	DGW: 66 (66.7); TPS: 81 (77.9)	DGW: 1; TPS: 0	DGW: 1; TPS: 2	DGW: 16 (16.2); TPS: 14 (13.5)	PD stent in 11 (DGW), 8.7 (TPBS)
Maeda et al[31], 2003	No cannulation after 10 minutes	Not defined	53	DGW: 64; RSG: 64	DGW: 26 (40); RSG: 29 (45.3)	DGW: 27; RSG: 26	DGW: 25 (92.6); RSG: 14 (53.8)			DGW: 0; RSG: 0
Yoo et al[32], 2013	No cannulation after 10 attempts	Not defined	71	DGW: 67 ± 11.5; TPS: 63.7 ± 16.5	DGW: 18 (52.9); TPS: 23 (62.2)	DGW: 34; TPS: 37	DGW: 27 (79.4); TPS: 29 (78.4)	DGW: 1 (2.9); TPS: 2 (5.4)	0/0	DGW: 13 (38.2); TPS: 4 (10.8)	No PD stent
Coté[33], 2012	No cannulation in 6 minutes or PD cannulation 3 times	Not defined	87	DGW: 58.1 ± 17.2; PDS: 57.4 ± 16.9	N/A	DGW: 42; PDS: 54	DGW: 16 (38.1); PDS: 26 (57.8)	DGW: 0; PDS: 0	DGW: 0; PDS: 0	DGW: 1; PDS: 3
Sugiyama et al[36], 2018	No cannulation within 15 minutes or PD cannulation > 3 times	3 years’ experience; > 300 ERCP per year	68	TPS: 69.8 ± 9.3; DGW: 67 ± 11.3	TPS: 18 (52.9); DGW: 19 (55.9)	TPS: 34; DGW: 34	TPS: 33 (94.1); DGW: 20 (58.8)	TPS: 0; DGW: 0	TPS: 0; DGW: 0	TPS: 1 (2.9); DGW: 1 (2.9)	PD stent placed in all
Zang et al[37], 2014	No cannulation within 10 minutes or PD cannulation up to 5 times	Not defined	149	TPS: 53.7 ± 9.2; NKS: 54.6 ± 10.8	TPS: 36 (49.3); NKS: 35 (46)	TPS: 73; NKS: 76	TPS: 70 (95.9); NKS: 64 (84.2)	TPS: 1 (1.4); NKS: 3 (3.9)	TPS: 0; NKS: 0	TPS: 6 (6.8); NKS: 5 (6.6)	No PD stent; All wire guided cannulations

DBC: Difficult biliary cannulation; PD: Pancreatic duct; PEP: Post endoscopic retrograde cholangiopancreatography pancreatitis; CBD: Common bile duct; WGC-PS: Wire-guided cannulation - pancreatic stent; DGW: Double guidewire; PD-1: Pancreatic duct guidewire; RSG: Repeated single guidewire cannulation; N/A: Not applicable; NKF: Needle knife fistulotomy; TPS: Trans-pancreatic sphincterotomy; NKS: Needle knife sphincterotomy; ERCP: Endoscopic retrograde cholangiopancreatography.

Trans- pancreatic sphincterotomy

This technique was first described by Goff[35] in 1995. The principal was to cut the septum between the PD and CBD, so as to gain entry into the terminal portion of the CBD. In cases of inadvertent PD cannulation, guidewire was first inserted into the PD to stabilize the sphincterotome. The sphincterotome was then directed toward 11 o’clock in order to cut the septum and gain entry into the CBD. More than half of the cutting wire was kept outside of the PD. Cuts were made in small increments of not more than 5 mm with positional readjustments, if required. Biliary sphincter was usually found either lateral or superior-lateral to the PD. Once the biliary sphincter or mucosa is identified, the sphincterotome is removed from the PD, and an attempt is made to cannulate the exposed bile duct directly. Successful CBD cannulation was achieved in 28 of 32 patients using this technique (87.5%). Four patients developed PEP (12.5%), while there were no other significant complications. The most worrisome complication with this technique is the occurrence of PEP. Placement of guidewire in the PD can cause mechanical injury, and thermal injury can occur during the septotomy. Although not described by Goff[35], it may be prudent to leave a guidewire in the PD and place a pancreatic stent, to prevent PEP.

Successful cannulation rate in various RCTs study TPS ranged from 77.9% to 95%[30,36,37]. PEP rates varied from 1%-13%. Incidence of PEP after using TPS technique depends on multiple factors including the time duration for which manipulation occurred at the papilla, contrast injection into the PD and placement of pancreatic stent. Preventive measures, including early switch to pre-cut techniques, exclusive use of guidewire cannulation and ensuring placement of pancreatic stent can decrease the incidence of PEP. There is a hypothetical risk of papillary stenosis and chronic pancreatitis after TPS. A retrospective study comparing long term outcomes of TPS and a control group of patients with conventional cannulation found no significant difference. Follow up data were available for more than four years after the index ERCP. There was no case of chronic pancreatitis or stenosis in the TPS group[38]. Another retrospective study compared the long- and short-term outcomes of patients (n = 125) who underwent ERCP using either TPS (n = 82) or NKPP (n = 43) technique. There was no case of chronic pancreatitis or papillary stenosis in either group on follow up of more than 6 months[39].

Table 2 describes RCTs which compare TPS with other techniques. Two meta-analysis comparing TPS to NK pre-cut techniques resulted in higher cannulation and lower bleeding rates in the TPS group. Odds of PEP and other complications were similar between both the groups[40,41]. Another meta-analysis compared TPS with NKPP, NKF and DGW individually. TPS showed superiority in cannulation rate compared to DGW [odds ratio (OR): 2.72] and NKPP (OR: 2.32) while success rate did not differ from NKF [OR: 1.38; 95% confidence interval: 0.32-5.96]. No difference was found between TPS vs DGW and TPS vs NKPP when comparing the incidence of PEP. Odds of PEP were significantly higher than NKF (OR: 4.62; 95% confidence interval: 1.36-15.72). Odds of bleeding and perforation were similar in each group[42].

PD STENT ASSISTED TECHNIQUES

Inadvertent PD cannulation increases the risk of PEP, and European Society of Gastrointestinal Endoscopy has included guidewire insertion into the PD more than one time in their definition of DBC[1]. Placement of a PD stent temporarily can alleviate the intra ductal pressures and has been proved to decrease the risk of PEP[43]. PD stent may also facilitate CBD cannulation by straightening the common channel, making it easier to find the direction needed for biliary cannulation.

Slivka[44] described the process of PD stenting followed by use of an ultra-tapered cannula for biliary cannulation. Coté et al[45] performed a retrospective cohort study which showed a successful cannulation rate of 93.4% when using the technique of wire guided cannulation over a PD stent. Retrospective data comparing DGW technique with wire guided cannulation-PD stent technique have shown similar canulation and PEP rates[46]. A prospective randomized study compared DGW cannulation technique to wire guided cannulation-PD stent technique noted a higher rate of successful cannulation in the DGW group. Rescue with pre-cut over PD stent was done more frequently in the wire guided cannulation over PD stent arm. Adverse events including PEP were minimal and similar in both groups[25].

Madácsy et al[47] first described the technique of combining the process of PD stent insertion and NK papillotomy. They hypothesized that early placement of the PD duct stent followed by NK papillotomy would not only enable successful biliary cannulation but also decreased the risk of PEP. They performed this procedure in 22 patients with sphincter of Oddi dysfunction. Successful biliary cannulation was achieved in 90% cases and there was no case of PEP. NK pre-cut papillotomy over PD stent (Figure 2) has been found to be more effective and safer (in terms of PEP) than free hand NK papillotomy[48].

Figure 2 Pre-cut sphincterotomy over pancreatic duct stent. A: Inadvertent pancreatic duct (PD) cannulation during conventional endoscopic retrograde cholangiopancreatography (guidewire entering PD) on fluoroscopic image; also noted are metallic clips near cystic duct stump; B: Endoscopic image showing PD stent in place and needle knife for performing fistulotomy over PD stent; C: Fistulotomy performed over the bulging papilla using needle knife sphincterotome to access the bile duct; D: Once adequate fistulotomy was achieved, standard sphincterotome was used to selectively canulate the bile duct; E: Once selective cannulation was achieved, a self-expanding metal stent was loaded over the guidewire; F: Self-expanding metal stent deployed in the biliary system with free flow of contrast; also noted is the PD stent in place.

EUS GUIDED RV TECHNIQUE

First case of EUS-guided RV technique was first described for selective PD cannulation in a patient with chronic pancreatitis[49]. The technique was then replicated by Mallery et al[50] for selective biliary cannulation as well. RV technique can be used for selective biliary cannulation in patients with an accessible papilla. RV means a pre decided meeting, all RV procedures involve passing the guidewire across the papilla which the endoscopist can then use to facilitate selective biliary cannulation at the site of the meeting i.e. the papilla.

EUS-RV technique involved direct puncture of the bile duct with a 19-G or 22-G under endosonographic vision. The size of the needle depends on the diameter of the duct at the puncture point and the choice of the endoscopist. While no definite cut off exists for minimal ductal dilatation, usually duct diameter of less than 3 mm are preferably punctured with a 22-G needle. One of the key drawbacks of the 22-G needle is that it allows only the passage of an 0.018’ guidewire and this may require changing to a 0.025’/0.035’ guidewire during the latter part of the procedure. Access can either be obtained at the level of intra-hepatic ducts from the gastro-esophageal junction or the antrum (Figure 3), or at the level of extra hepatic bile duct from duodenal bulb or the second part of the duodenum (Figure 4). The most direct route from needle puncture site to the papilla is attained from the second part of duodenum. However, the scope position may be unstable. Advantages of puncturing from the duodenum is that the bile duct is retroperitoneal and risk of biliary peritonitis is minimized. Scope position should be manipulated such that the needle tip faces towards the papilla on fluoroscopy[51]. After access has been confirmed by bile aspiration and contrast injection, a guidewire is passed into the biliary system and manipulated such that the guidewire crosses the papilla in antegrade fashion and forms a loop in the duodenal lumen. Either a 0.025 inch or 0.035-inch steerable hydrophilic guidewire should be used. Crossing of the guidewire across the papilla is the rate limiting step of this technique and determined the overall success rates. Likelihood of successful crossing across the papilla depends on the distance from the papilla and the direction of the needle, both of which are influenced by the EUS scope position. Once the wire has crossed into the duodenal lumen, the EUS scope and needle are removed under fluoroscopic vision, leaving the wire in situ. The duodenoscope is then introduced. Selective biliary cannulation can then be achieved in two ways. The RV guidewire may be grasped using either a snare or foreign body forceps and pulled back into the working channel of the duodenoscope. A sphincterotome or cannula can then be loaded onto this wire for selective biliary cannulation. Alternatively, wire guided cannulation may be attempted adjacent to the RV wire as the RV wire delineates the trajectory of the CBD. Once selective biliary cannulation has been achieved using a guidewire, the RV wire can be removed.

Figure 3 Conventional endoscopic ultrasound guided rendezvous technique (intrahepatic route): A: Endoscopic ultrasound (EUS) image showing dilated intrahepatic biliary radicle and puncture made using 19-G EUS-fine needle aspiration needle; B: After bile aspiration and confirming bile duct puncture, guidewire passed across the papilla into the duodenum as shown in the fluoroscopic image; C: After coiling the guidewire in the duodenum, the EUS scope was exchanged over the guidewire; D: Endoscopic image of the papilla showing a straight plastic stent placed in pancreatic duct (at the time of difficult biliary cannulation during conventional endoscopic retrograde cholangiopancreatography) and the guidewire in the bile duct passed during EUS-rendezvous procedure; E: Selective cannulation of bile duct performed using a sphincterotome by the side of the guidewire; F: Fluoroscopic image showing deep cannulation of the bile duct during endoscopic retrograde cholangiopancreatography procedure; also noted is the guidewire within the bile duct (placed during EUS-rendezvous procedure) with pancreatic duct stent in place.

Figure 4 Conventional endoscopic ultrasound guided rendezvous technique (extrahepatic route) being performed in a case of acute biliary pancreatitis (percutaneous catheter drain in situ) diagnosed with choledocholithiasis. A: Endoscopic ultrasound (EUS) scope positioned in short scope position in D1-D2 junction; B: After common bile duct puncture (access diameter 43 mm) using 19 G EUS- fine needle aspiration needle, bile aspiration was done to confirm position and thereafter, contrast was injected to delineate the bile duct; C: Guidewire was passed across the papilla and coiled in the duodenum; D: EUS scope was exchanged with guidewire in place; E: Subsequent endoscopic retrograde cholangiopancreatography was performed using a duodenoscope and successful biliary cannulation was achieved along the side of the exiting guidewire; F: After stone extraction using balloon sweeps, common bile duct stent (7 French × 7 cm double pigtail stent) was placed.

In a prospective study by Iwashita et al[52], successful biliary cannulation using EUS-RV was noted in sixteen out of twenty patients (80%). Guidewire manipulation was successful in 100% of the patients from the second part of the duodenum and 66.7% of the patients from the duodenal bulb or gastroesophageal junction. Two patients (10%) developed mild pancreatitis and 1 (5%) developed a liver hematoma. However, there were no adverse events related to bile leak. Two RCTs comparing NKF with EUS-RV found the rate of successful cannulation is similar between the two groups[53,54], however, PEP was lower in EUS-RV group in the RCT by Dhir et al[54]. Table 3 describes the two RCTs as discussed previously. A meta-analysis of 12 studies found the clinical success rate of EUS-RV ERCP to be 80.8% with an overall complication rate of 14%[55,56].

Table 3 Studies (randomized controlled trials) comparing endoscopic ultrasound guided rendezvous technique with needle knife precut papillotomy, n (%)/mean ± SD.

Ref.	Definition of DBC	Endoscopist experience	Sample size	Baseline characteristics; age, years	Baseline characteristics; gender (male %)	Groups (n)	Successful CBD cannulation	Bleeding	Perforation	PEP	Comments
Choudhury et al[53], 2024	No cannulation after 5 minutes; > 5 contacts with papilla; > 1 unintended PD cannulation	Well versed with both ERCP and EUS-RV	100	EUS-RV: 50.52 (12.8); NKPP: 52.96 (15.7)	EUS-RV: 15 (30); NKPP: 13 (26)	EUS-RV: 50; NKPP: 50	EUS-RV: 92; NKPP: 90	0	0	EUS-RV: 5 (10); NKPP: 5 (10)
Dhir et al[54], 2025	No cannulation after 5 attempts, time > 5 minutes. All cases of distal biliary malignant obstruction	Not defined	208	NKPP: 64 (34-90); EUS-RV: 67 (42-84)	NKPP: 65 (62.2) EUS-RV: 57 (54.8)	NKPP: 104; EUS-RV: 104	NKPP: 97 (93.3); EUS-RV: 101 (97.1)	NKPP: 3 (2.9); EUS-RV: 2 (1.9)	NKPP: 0; EUS-RV: 0	NKPP: 9 (8.7); EUS-RV: 2 (1.9)	2 bile duct leak (1.9) in EUS-RV group

DBC: Difficult biliary cannulation; CBD: Common bile duct; PEP: Post endoscopic retrograde cholangiopancreatography pancreatitis; PD: Pancreatic duct; ERCP: Endoscopic retrograde cholangiopancreatography; NKPP: Needle knife precut papillotomy; EUS-RV: Endoscopic ultrasound guided rendezvous.

Percutaneous trans-hepatic biliary drainage and surgical RV techniques

Non endoscopic approaches to patients with failed biliary cannulation can either be percutaneous or surgical. Both approaches involve gaining antegrade access to the bile duct. Percutaneous trans-hepatic biliary drainage (PTBD) is a time tested and valuable technique for biliary drainage in cases where endoscopic biliary cannulation fails or is not feasible. PTBD involved puncture of the intra hepatic biliary radicles under ultrasound vision using a sheathed needle[57]. In the early days of interventional radiology, catheters were left above the obstruction to drain the bile externally. While this led to an effective decompression if the biliary system, presence of an external drainage system was found to be cumbersome. Risk of occlusion or slippage of the percutaneous catheter further increased the risk of an external biliary fistula formation and cholangitis[58]. However, with development of multi-side hole catheters it became possible to manipulate guidewires and catheter across the obstruction and papilla and into the duodenum. In this manner, trans papillary drainage could be achieved even with an initial percutaneous access. Complications of PTBD, though rare, include acute hemorrhagic complications, trans pleural punctures leading to pneumothorax and local skin complications. A hybrid PTBD-endoscopy approach involves using the PTBD guidewire and catheter to RV with the duodenoscope at the papilla with subsequent retrograde biliary cannulation (PTBD-RV)[59]. After needle puncture of the intra hepatic bile duct, the needle is exchanged for an introducer sheath through which the guide wire is manipulated across the papilla. In case of difficulty in traversing across the papilla, antegrade papillary balloon dilatation can be done using 5-7 Fr balloons. Thereafter, the patient is turned to a left lateral or semi prone position and a duodenoscope is introduced. The endoscopist can then cannulate the biliary system using various RV techniques[59,60]. The endoscopist can either cannulate adjacent to the protruding wire (parallel cannulation technique)[61] or they can pull the guidewire into the accessory channel using a snare or a Dormia basket. Sphincterotome can then be loaded over the wire for the biliary cannulation push pull technique[62]. Push pull technique is most useful in patients with tight biliary strictures. However, there is risk of injury to the hepatic parenchyma and capsule during guidewire manipulation. PTBD-RV can be performed as a single stage or two stage procedure[63]. When the two stage procedure is performed, initially the catheter is left proximal to the obstruction and RV cannulation takes place at the second stage. High technical success rate has been observed with both PTBD and PTBD-RV (89% vs 80% respectively). Fewer complications were noted with PTBD-RV. The cholangioscope can also be passed through the percutaneous catheter and assist in guidewire negotiation and RV procedure in patients with distorted anatomy[64].

In an open label multi-center study, outcomes of PTBD vs EUS-biliary drainage were compared in cases with malignant hilar obstruction[65]. While technical, clinical success rates and complication rates were similar between the two groups. Incidence of recurrent obstruction was significantly higher in those who underwent PTBD. When comparing outcomes of PTBD vs EUS biliary interventions in distal malignant obstruction, EUS guided procedures had fewer complications, shorter length of hospital stay and lower mortality rates[65,66]. A meta-analysis of 13 studies compared the outcomes of EUS-RV with PTBD-RV. While success rates and complication rates were similar in both malignant and benign cases, technical success was higher with PTBD-RV in cases with altered surgical anatomy[67].

Treatment of patients with cholelithiasis and choledocholithiasis entails a cholecystectomy with a pre or post op ERCP. This two-step approach, while convenient, increases the overall cost and length of hospital stay. Intra-op ERCP and RV procedures may be useful in this category of patients, especially in those with suspected DBC. During the intra op-RV procedure, the surgeon inserts a 5 Fr catheter through the cystic duct under laparoscopic vision. A guidewire is passed through the catheter into the bile duct and negotiated across the papilla into the duodenum. Remainder of the RV steps are similar to those used in PTBD-RV or the EUS-RV procedures. The endoscopist passes the duodenoscope via mouth and cannulation is attempted either adjacent to the guidewire or the guidewire is caught using a Dormia basket and withdrawn into the accessory channel. Thereafter a papillotome is introduced over the guidewire to perform ERCP. The PEP rate of 2% was noted with RV procedure compared to intra op ERCP (13%)[68]. A meta-analysis which compared two step procedures with single step procedures found lap cholecystectomy and intra op ERCP to be most likely to have favorable success and safety outcomes[69].

ERCP IN ALTERED ANATOMY

Surgically altered anatomy can be divided into two types. Type I where duodenum is still in continuity with the stomach and a conventional duodenoscope may be used. Examples include Billroth type 1 gastrectomy and sleeve gastrectomy. In type II surgically altered anatomy, stomach remnant or the stomach is not in continuity with the duodenum. It is difficult to reach the papilla using a conventional duodenoscope. Examples include surgeries like Billroth II gastrectomy, Roux-en-Y gastric bypass (RYGB) or any other surgery involving creation of a Roux-en-Y diversion. Using the conventional duodenoscope, the papilla was accessible in only one third of the patients with RYGB and two third of the patients with Billroth type II surgery[70]. Single balloon or double balloon enteroscopes have a sufficient length to reach the papilla with ease[71]. However, due to length of the enteroscope, conventional ERCP accessories may not be compatible. Furthermore, lack of an elevator makes the process technically challenging. Short length enteroscopes were developed for the purpose of enteroscopy guided ERCP. Another method to do ERCP in altered anatomy is the use of a pediatric colonoscope[72]. Underwater immersion technique can improve stabilization and visualization of the intestinal structures[73]. Age > 75 years, Roux-en-Y construction, pancreatic indication and malignant indication are associated with higher chances of technical failure[74].

Another approach in patients with surgically altered anatomy is use of EUS guided procedures. EUS-RV procedures can be done in patients with type I surgically altered anatomy with ease. It can also be attempted in patients with type II anatomy; however, papilla would still need to be approached using enteroscopes. In patients with malignant obstruction or those in which papilla is not accessible even with enteroscopes, antegrade or transluminal approaches (including EUS-hepaticogastrostomy and EUS-hepaticojejunostomy) can be attempted[75]. A multicenter prospective study of patients with surgically altered anatomy and malignant biliary obstruction found EUS-biliary drainage to have higher success rates and significantly shorter procedure time than enteroscopy assisted ERCP[76].

EUS-guided trans-gastric ERCP is another newer technique for performing ERCP in patients who underwent RYGB[77] (Figures 5 and 6). The technique involves two steps. First, EUS guided access is gained into the excluded stomach. A temporary trans-gastric fistula is created between the gastric pouch or the proximal jejunum and the excluded part of the stomach. This is done by placing a lumen apposing metal stent (LAMS) between the two luminal organs under EUS vision. LAMS creates a stable conduit to the excluded stomach. LAMS can be dilated to allow entry of a conventional duodenoscope. Thereafter ERCP can be done in the conventional fashion. LAMS can then be removed after 4 weeks. The fistula track either closes spontaneously or can be closed using techniques like over the scope clip or endoscopic suturing. A multicenter retrospective study of 178 patients documented a 98% success rate in LAMS placement and successful biliary cannulation in 100% of the cases in whom LAMS could be placed[78]. Adverse events included those from LAMS deployment, ERCP and failure to close the fistula tract. In the above study, perforation occurred in 6 patients (3%), LAMS migration in 2 patients (1.1%), mis-deployment in 9 patients (5%). Nine patients had persistent fistula 8 weeks after the procedure.

Figure 5 Endoscopic ultrasound-directed trans-gastric endoscopic retrograde cholangiopancreatography for management of biliary obstruction in altered anatomy. A: Endoscopic ultrasound (EUS)-directed trans-gastric endoscopic retrograde cholangiopancreatography being performed in a patient for the management of choledocholithiasis in a Roux-en-Y gastric bypass anatomy; Using linear echoendoscope, under EUS and fluoroscopic guidance, the excluded stomach was located endosonographically from the remnant gastric pouch; B: Once an optimal position was confirmed, the remnant stomach was punctured using a 19-G EUS-fine needle aspiration (FNA) (EZ shot 3 plus; Olympus, Japan) needle; C: Contrast along with 120 mL of water was injected using the FNA needle to confirm the position within and distend the excluded stomach; D: EUS image shows the distended stomach with FNA needle in-situ; E: The fistula tract was created using the freehand technique, by electrocautery-enhanced lumen apposing metal stent (size 20 × 10 mm; Hot-AXIOS; Boston Scientific, MA, United States) where the distal end was deployed in the excluded stomach and the proximal flange into the remnant gastric pouch; F: Fluoroscopic image shows the presence of lumen apposing metal stent in-situ creating a trans-mural tract.

Figure 6 Endoscopic ultrasound-directed trans-gastric endoscopic retrograde cholangiopancreatography for management of biliary obstruction in altered anatomy. A: Endoscopic image shows the presence of lumen apposing metal stent (LAMS) in-situ in the remnant gastric pouch; this is the route which is used through which the duodenoscope is passed to each of the papilla; B: Once papilla is reached, selective cannulation of bile duct is performed using a standard sphincterotome; C: Fluoroscopic image shows the guidewire in the common bile duct (CBD), following which contrast was given and cholangiogram was taken. Cholangiogram shows presence of dilated intrahepatic biliary radicles with a filling defect in CBD (likely stone); also noted are presence of metallic clips near cystic duct (indication of past laparoscopic cholecystectomy); D: Balloon sweeps were taken to clear the CBD of the stones or sludge; E: Check cholangiogram shows no filling defect in CBD; yellow box denoted the presence of LAMS through which the duodenoscope has passed; after completion of endoscopic retrograde cholangiopancreatography, LAMS was removed in same session; F: Follow up endoscopy after a month revealed persistent gastro-gastrostomy fistula; G: Margins were made raw by argon plasma coagulation; H: The fistula was closed using an 11/6t over-the-scope clip.

APPROACH TO PATIENTS WITH DBC

Success of biliary cannulation depends on multiple factors including duodenoscope position, papilla morphology, endoscopist’s experience and the type of technique used. The technique to be chosen for any particular case should be decided based on papilla’s morphology, endoscopist experience and whether there was inadvertent PD cannulation. Various classifications exist to identify the morphology of papilla and associated risks. Haraldsson et al[79] classified the papilla morphology into four types: (1): Type 1-regular papilla; (2) Type 2-small flat papilla, diameter less than 3 mm; (3) Type 3-protruding papilla; and (4) Type 4-creased papilla. A retrospective study identified that advanced cannulation techniques are required more commonly for type 2 and type 4 papilla, while pre-cut sphincterotomy may be most effective in type 3 morphology[78]. PEP was most commonly associated with type 2 morphology[80]. Horiuchi classified the papilla morphology as small, large and swollen[81]. They proposed the use of TPS for small, NKPP for large and NKF for swollen papilla. Success cannulation with each technique was 96%, 90% and 100%, respectively.

A network meta-analysis of 17 RCTs compared the efficacy of different methods for DBC[82]. Higher odds for successful biliary cannulation were noted with TPS compared to standard cannulation, pancreatic guidewire assisted cannulation, early NK and PD stent assisted techniques. On network model development, highest scores were seen with TPS followed by early NK techniques. PEP rates were also lowest with these two techniques.

Depending on whether PD cannulation occurred or not, technique for DBC can be decided. In cases with pancreatic guide-wire insertion and a type 1 or 4 morphology-TPS or DGW technique may be attempted. In type 2 morphology, the double guidewire technique may be preferred. In type 3 morphology, PD stent insertion and NKPP over the stent may lead to better outcomes. In all cases with PD guidewire insertion, PD stent placement is recommended to decrease the risk of PEP. In cases with no guidewire in the PD, NKF may be attempted in type 1 or type 3 morphology. NK precut papillotomy may be tried in type 4 morphology. In type 2 morphology, the risk of perforation is high and any NK technique should be used with caution. EUS guided RV can be done irrespective of papilla morphology but it requires considerable expertise. If there is failure to gain access despite these advanced techniques, a re-attempt should be made after 48-72 hours. In patients with failed NKPP, a repeat attempt after 72 hours led to a successful cannulation in 76.8% of the patients[83]. Ultimately, if all techniques fail and a patient needs urgent biliary drainage, PTBD acts as an important rescue procedure.

Nevertheless, while these various options are available, the choice and the sequence adopted will depend also on expertise of the operator, the availability of the facility at the performing center and the equipment thereof. Moreover, the learning curve for these various salvage techniques differ and the impact of the center’s case volume will have an impact on that. For example, the facility of EUS-RV may not be available at all centers and even if the facility is available, the expertise might be lacking. Thus, the algorithm for a center in adopting the various salvage techniques for a particular case of DBC will depend on these additional factors as well. A pragmatic algorithmic approach, considering that all the facilities and expertise are available, has been outlined in Figure 7.

Figure 7 Algorithm for approach to a case of difficult biliary cannulation during endoscopic retrograde cholangiopancreatography. PD: Pancreatic duct; NK: Needle knife; PDS: Pancreatic duct stent; TPS: Transpancreatic sphincterotomy; DGW: Double guidewire; NKF: Needle knife fistulotomy; NKPP: Needle knife precut papillotomy; EUS-RV: Endoscopic ultrasound guided rendezvous technique; PTBD: Percutaneous trans-hepatic biliary drainage.

CONCLUSION

DBC remains a key obstacle during ERCP, demanding a stepwise approach to optimize outcomes and minimize adverse events. Early recognition of difficult anatomy or technique failure allows timely transition to alternative strategies such as DGW, TPS, NK or EUS-RV. Operator experience, hospital volume, patient selection, and use of prophylactic measures (such as PD stenting, rectal non-steroidal anti-inflammatory drugs, intravenous hydration) to prevent PEP are essential. Advances in technology and resident training continues to improve technical success and safety of the procedure. Ultimately, managing DBC requires a delicate balance between persistence and prudence, with a low threshold for expert referral when necessary to ensure appropriate patient centered care.