Endoscopic ultrasonography-guided biliary drainage (EUS-BD) is gaining popularity as the procedure of choice over percutaneous methods in patients with failed biliary cannulation in endoscopic retrograde cholangiopancreatography.
The different approaches in EUS-BD are associated with different outcomes and the choice should be tailored to individual patient needs according to an algorithm.
EUS-guided pancreatic duct drainage is a safe and feasible option for pancreatic ductal drainage in patients with symptomatic obstructed pancreatic duct.
EUS-guided transmural gallbladder drainage is becoming the choice of procedure over percutaneous methods to drain the gallbladder in patients suffering from acute cholecystitis and who are unfit for cholecystectomy.
Advanced techniques and dedicated devices are required for the above procedures to achieve superior efficacy and good safety profile of the procedures.
Endoscopic retrograde cholangiopancreatography (ERCP) is currently an indispensable tool for achieving endoscopic biliary and pancreatic duct drainage. Successful deep cannulation of the biliary tree can be achieved in ≥90% of ERCPs performed by the experienced endoscopist. Failure of cannulation could occur in patients with unusual anatomy or obstructing tumors, or when the papilla is inaccessible due to surgically altered anatomy. In patients with failed ERCP, biliary drainage could be achieved by percutaneous or surgical means, whereas pancreatic duct drainage could only be achieved by surgery. Percutaneous biliary drainage is associated with technical success rates of 77% to 100% and adverse events rates of 6% to 31%. However, it is often less preferred by patients as the external drainage tube causes inconvenience and tube-related problems. Surgical biliary drainage is associated with lower rates of recurrent biliary obstruction, but the invasive nature of the procedure causes more adverse events resulting in a longer hospital stay.
Recently, endoscopic ultrasonography (EUS)-guided biliary and pancreatic duct drainage has gained popularity as an alternative approach to achieve bile or pancreatic duct drainage. The approach could achieve internal biliary and pancreatic duct drainage, with multiple options of access routes depending upon the anatomy and level of obstruction at the same session of a failed ERCP. This chapter will provide an overview on the various approaches, techniques and accessories, advantages and disadvantages, the reported outcomes, and the adverse events of these techniques.
Endoscopic Ultrasonography-Guided Biliary Drainage
EUS-guided biliary interventions comprise of a group of procedures where the echoendoscope is used for bile duct access, drainage, and stone extraction. The types of procedures are shown in Fig. 24.1 . EUS-rendezvous ERCP (EUS-Rv biliary) is a type of access procedure, where the role of EUS is to provide bile duct access and insertion of a guidewire to guide subsequent cannulation during ERCP. EUS-guided biliary drainage (EUS-BD) procedures are divided into transluminal and antegrade procedures depending on whether a neo-transluminal fistula is created for drainage (transluminal), or if drainage is done in an antegrade fashion from the intrahepatic ducts toward the distal bile duct in the native biliary system (antegrade). Transluminal drainage could be further classified by the anatomical site of drainage into choledochoduodenostomy (CDS) and hepatico-gastrostomy (HGS). A CDS involves EUS-guided creation of a transmural fistula between the first part of the duodenum and the distal bile duct followed by placement of a stent. EUS-HGS involves EUS-guided creation of a transmural fistula between the left intrahepatic ducts and the stomach followed by placement of a stent. Antegrade procedures could be further classified by the location of the distal end of the stent into trans-anastomoic, suprapapillary, and trans-papillary procedures. It involves EUS-guided transhepatic puncturing of the intrahepatic ducts, followed by antegrade placement of a stent. EUS-guided antegrade stone extraction is used for stone extraction when the papilla is not accessible, often in patients with prior gastric bypass and a long afferent limb. It involves EUS-guided antegrade advancement of stones across the ampulla using a balloon catheter. The use of a standard nomenclature is important, as individual procedures differ markedly in efficacy and risk profiles, and their outcomes should not be analyzed together.
Indications of Endoscopic Ultrasonography-Guided Biliary Drainage
The current indications of EUS-BD are listed in Table 24.1 . The most common indication of EUS-BD is failed deep cannulation of the bile duct or an inaccessible papilla. The decision to perform EUS-BD in the event of failed ERCP should depend on the institution’s success rates in achieving biliary cannulation with advanced ERCP techniques and the availability of endosonographers experienced in EUS-BD. In the event of difficult cannulation, advanced ERCP techniques should achieve cannulation in 73.4% to 100% of the patients. Thus the use of EUS-BD as a salvage for failed ERCP should be uncommon.
|Indications of EUS-Guided Biliary Drainage|
|Indications of EUS-Guided Pancreatic Duct Drainage|
|Indications of EUS-Guided Gallbladder Drainage|
Outcomes of the Endoscopic Ultrasonography-Guided Biliary Drainage Procedures
Endoscopic Ultrasonography-Rendezvous Endoscopic Retrograde Cholangiopancreatography (Biliary)
EUS-Rv is favored by many endoscopists over the transluminal techniques as it avoids the formation of a permanent bilio-enteric fistula and the need to dilate the fistulous tract, which may lead to an increased risk of adverse events. However, the EUS-Rv approach may not be conceivable if the guidewire cannot be advanced through the site of obstruction and papilla, often due to severe ductal dilation, angulation, or a tight biliary/anastomotic stricture. The outcomes of studies with more than 20 patients that received EUS-Rv are shown in Table 24.2 . The results from comparative studies are limited and mostly included a small number of patients. A retrospective study compared EUS-Rv with precut sphincterotomy after failed cannulation for benign and malignant conditions. Patients with failed cannulation after precut sphincterotomy received a second ERCP 72 hours later. Both methods were comparable in the overall rates of cannulation, but the EUS-Rv group had higher cannulation rates in the first session. There were no differences in the rate of adverse events, but the groups differed in risk profiles, with more patients suffering from pancreatitis and bleeding in the precut group, and more periductal contrast leak in the EUS-Rv group. In another study, two ERCP cohorts comprising more than 1000 patients in each group were compared. EUS-BD methods were employed in the event of failed cannulation in one group. The failure rate when both precut and EUS-BD methods were available was significantly lower than the group with only ERCP available (1% vs. 3.6%, P < .001). The success rate of EUS-BD was also significantly higher than for precut sphincterotomy (95.1% vs. 75.3%, P < .001), mainly due to superiority in patients with malignant obstruction (93.5% vs. 64%, P < .001). Khashab et al. also compared patients that underwent EUS-BD with the rendezvous technique (13 patients) versus the transluminal technique (20 patients). There were no differences in technical and clinical success, procedural time, length of hospital stay, and adverse events. A study then compared EUS-Rv with percutaneous biliary drainage in patients with malignant distal biliary obstruction. A lower success rate in obtaining biliary drainage was observed in the EUS-Rv group (76% vs. 100%, P = .002). However, the length of hospital stay was shorter ( P = .02) and the need for repeated biliary interventions was lower ( P = .001) in the EUS-Rv group.
|Author||Year||Patients||Technical Success||Extrahepatic||Transhepatic||Adverse Events|
|Kahaleh et al.||2006||23||65%||4/5||11/18||17%|
|Shah et al.||2012||50||75%||—||—||12%|
|Iwashita et al.||2012||40||73%||25/31||4/9||13%|
|Dhir et al.||2012||58||98%||57/58||0||3%|
|Vila et al.||2012||60||68%||—||—||22%|
|Park et al.||2013||20||80%||3/6||13/14||0%|
|Dhir et al.||2013||35||97%||18/18||16/17||23%|
|Iwashita et al.||2016||20||80%||13/15||3/4||15%|
|Tang et al.||2016||25||80%||20/24||1/1||—|
|Lee et al.||2017||50||94%||5/5||42/45||—|
|Bill et al.||2016||20||76%||—||—||28%|
Endoscopic Ultrasonography-Guided Antegrade Stenting
The outcomes of EUS-guided antegrade stenting (EUS-AG) are less well reported than other EUS-BD procedures. EUS-AG is usually chosen in patients where the papilla is not accessible and the guidewire is successful in traversing the stricture. The studies that reported the individual outcomes for EUS-AG are shown in Table 24.3 . Overall, EUS-AG seems to be associated with high technical success rates and an acceptable risk of adverse events. However, EUS-AG also suffers from difficulties of guidewire manipulation as in EUS-Rv, thus some endoscopist may prefer transluminal techniques over EUS-AG.
|Author||Year||Patients||Technical Success||Adverse Events|
|Park et al.||2013||14||60%||14%|
|Ogura et al.||2014||12||100%||8.3%|
|Dhir et al.||2014||25||92%||32%|
|Iwashita et al.||2017||20||95%||20%|
Endoscopic Ultrasonography-Guided Choledochoduodenostomy and Hepaticogastrostomy
EUS-CDS and HGS are both transmural techniques that achieve biliary drainage. However, there are several important differences between the two procedures. In making a CDS, a fistula is created between the common bile duct and the duodenum. The common bile duct is located directly behind the first part of the duodenum and is relatively fixed with minimal respiratory influence. Thus the risk of the organs separating after creation of the fistula is low, and the risk of stent migration is uncommon. In EUS-HGS, needle puncture through the thicker gastric wall and a few centimeters of hepatic parenchyma with greater tissue resistance is required. This makes puncture and stent deployment more challenging. Furthermore, the stomach is regularly undergoing peristalsis and the liver moves during respiration. The stent placed is more prone to the risk of migration resulting in bilomas or free perforation. These complications may be fatal, and attention to technical details is therefore paramount. In the event of misdeployment, salvage stent placement may be impossible. In addition, because HGS involves transhepatic puncturing of the bile ducts, the procedure is susceptible to similar risks as seen during percutaneous transhepatic puncturing of the ducts. These procedural characteristics are responsible for the differences in results seen in both procedures.
The outcomes of studies with more than 20 patients are shown in Table 24.4 . The overall clinical success rates of both procedures in expert hands were 63.2% to 100%.
|Author||Year||Patients||Type of Procedure||Clinical Success (%)||30-Day Adverse Events (%)||Mean Duration of Stent Patency (Days)|
|Park et al.||2011||57||CDS 26 |
|Villa et al.||2012||65||CDS 26 |
|Dhir et al.||2014||45||—||93.7||22.9||—|
|Kawakubo et al.||2014||64||CDS 44 |
|Song et al.||2014||27||CDS 17 |
|Dhir et al.||2015||104||CDS 68||98.3||8.65||—|
|Poincloux et al.||2015||92||CDS 26 |
|Ogura et al. , a||2014||51||All HGS||100||—||202|
|Umeda et al.||2015||23||All HGS||100||4.3||120|
|Cho et al.||2016||54||CDS 33 |
|Kunda et al.||2016||57||All CDS||94.7||7||—|
|Khashab et al.||2013||20||CDS 15 |
|Khashab et al.||2016||121||CDS 60 |
|Kawakubo et al.||2016||26||All CDS||96.2||26.9||—|
|Ogura et al.||2016||39||CDS 13 |
The rate of adverse events was between 4.3% and 26.9%. The mean duration of stent patency was 37 to 329 days. The cause of stent dysfunction was mainly due to stones or sludge. The difference in outcomes of EUS-CDS and HGS are procedure specific.
In comparing EUS-CDS and HGS, the results from a few comparative studies are available. In a randomized trial, 49 patients with unresectable distal malignant biliary obstruction and failed ERCP received EUS-CDS or EUS-HGS. The technical success rates were comparable among the procedures (91% vs. 96%, respectively, P = .61). Clinical success was lower in the EUS-CDS group but did not reach statistical significance (77% vs. 91%, respectively, P = .23). HGS was associated with a higher risk of adverse events but it did not reach statistical significance (12.5% vs. 20%, respectively, P = .729).
In another study, a total of 121 patients underwent EUS-BD (CDS 60, HGS 61). The technical and clinical success rates were comparable among the two procedures ( P = .75 and P = .64, respectively). Adverse events were more common in the EUS-HGS group (19.67% vs. 13.3%; P = .37). Both plastic stenting (odds ratio [OR] 4.95; 95% confidence interval [CI]: 1.41 to 17.38; P = .01) and the use of noncoaxial electrocautery (OR 3.95; 95% CI: 1.16 to 13.40; P = .03) were independently associated with adverse events. The length of hospital stay was significantly shorter in the CDS group (5.6 days vs. 12.7 days; P < .001). The 1-year stent patency probability was greater in the EUS-CDS group (OR 0.98; 95% CI: 0.76 to 0.96 vs. OR 0.60; 95% CI: 0.35 to 0.78) but overall patency was not significantly different. Park et al. also studied predictors of adverse events in 57 patients who underwent EUS-CDS or HGS. Similar to the above studies, Park’s study showed no difference in technical success, clinical success, or rates of adverse events between two groups. However, the use of noncoaxial electrocautery (needle knife) was independently associated with the occurrence of adverse events (OR 12.4; P = .01).
Ogura et al. then compared EUS-CDS and HGS in patients with concomitant duodenal and biliary obstruction in a randomized study. No difference in technical success, clinical success, and adverse events rates were observed. However, EUS-CDS was associated with a significantly shorter duration of stent patency in an obstructed duodenum (43 days vs. 133 days; P = .05).
Khan et al. conducted a systematic review and meta-analysis of studies reporting on the outcomes of EUS-BD. Seven studies were included, and overall there was no difference in technical success between EUS-CDS and HGS (OR 1.32; P = .56). Six studies described postprocedure adverse events based on the method of drainage. EUS-CDS appeared to be significantly safer to HGS with a pooled OR of 0.40 ( P = .02).
In view of the above results, it could be concluded that both EUS-CDS and HGS are effective and safe techniques for the treatment of distal biliary obstruction after failed ERCP. However, EUS-CDS appears to be associated with a shorter hospital stay, improved stent patency, and fewer adverse events. In addition, metallic stents should be placed whenever feasible and noncoaxial electrocautery should be avoided when possible. In the presence of duodenal obstruction, EUS-HGS may be the preferred procedure as it is associated with a longer stent patency, whereas a CDS may be prone to restenosis due to tumor ingrowth or overgrowth.
Comparison of Endoscopic Ultrasonography-Guided Biliary Drainage With Percutaneous Transhepatic Biliary Drainage
Three randomized studies (one available in abstract only) have compared EUS-BD and percutaneous transhepatic biliary drainage (PTBD; Table 24.5 ). All have shown equivalent success rates. The adverse events rate and the need for reinterventions in two studies were significantly lower in the EUS-BD group. A meta-analysis then included nine comparative studies and 483 patients. There was no difference in technical success between two procedures (OR = 1.78; 95% CI: 0.69 to 4.59); however, EUS-BD was associated with better clinical success (OR = 0.45; 95% CI: 0.23 to 0.89), fewer postprocedure adverse events (OR = 0.23; 95% CI: 0.12 to 0.47), and lower rate of reintervention (OR = 0.13; 95% CI: 0.07 to 0.24). There was no difference in length of hospital stay after the procedures. Thus EUS-BD should be preferred over PTBD in the event of failed ERCP.
|Author||Year||Patients||Technical Success (%)||Clinical Success (%)||Adverse Events (%)||Reinterventions (%)|
|Artifon et al.||2012||EUS 13 |
|P value = NS|
|Lee et al.||2016||EUS 34 |
|P value = 0.022||25 |
|P value = 0.022|
|Giovannini et al.||2015||EUS 20 |
Pros and Cons of Each Procedure
The pros and cons of each procedure are shown in Table 24.6 . The considerations on planning the appropriate procedure include the difficulty of needle puncture, wire manipulation, dilation of the fistula tract, and stent placement. EUS-Rv is mainly an access procedure and usually does not involve fistula dilation. The risk of the procedure is mainly a result of the difficult ERCP, and there is limited added risk from the EUS intervention. However, the procedure is associated with several potential disadvantages. First, rendezvous completion is successful in 65% to 80% of the cases in most series and requires an accessible papilla, which may not be possible in patients with surgical upper gastrointestinal (GI) anatomy or gastric outlet obstruction (GOO). A second shortcoming is the need to exchange the echoendoscope for a duodenoscope, during which time there may be inadvertent displacement of the guidewire. A third limitation is the prolonged procedural times as a result of guidewire manipulation through the site of obstruction and ampulla, the need to exchange the echoendoscope for a duodenoscope, and the subsequent retrograde biliary interventions. A final shortcoming of EUS-Rv is the risk of acute pancreatitis due to manipulation of the papilla.
|Needle Puncture||Guidewire Manipulation||Tract Dilation||Ease of Stent Insertion||Stent Patency|
|EUS-Rv||Several options depending on the site obstruction and ductal dilation||Requires manipulation across the stricture and papilla||Usually not required||Similar to ERCP||Similar to ERCP|
|EUS-AG||Requires dilation of the intrahepatic ducts||Requires manipulation across the stricture and papilla||Required||Relatively straight forward||Similar to ERCP|
|EUS-CDS||Not possible in patients with duodenal invasion||Requires deep insertion into the intrahepatic ducts||Required||Similar to ERCP||6–12 months|
|EUS-HGS||Requires dilation of the intrahepatic ducts||Requires deep insertion into the intrahepatic ducts||Required||Technically demanding||3–4 months|
EUS-AG requires a dilated intrahepatic duct for needle puncture. The guidewire then needs to be negotiated across the site of obstruction for stent placement. Therefore it is susceptible to problems of guidewire manipulation as in the EUS-Rv procedure. However, if the guidewire is successful in traversing the stricture, the placement of the stent is relatively straightforward and is associated with a low risk of adverse events. There is a potential of bile leak from the access site, but the risk is low if drainage from the stent is adequate.
EUS-CDS and HGS are both transmural techniques and there are several advantages. These techniques are associated with no risk of pancreatitis as the papilla is not manipulated. The risk of tumor in-growth into the stent is also low as the stent placed does not traverse the tumor. Ductal puncture and access is usually easy in the presence of a dilated duct. However, the creation of the fistula tract requires multiple exchanges of instruments, and this may risk bile leak during the procedure. Furthermore, the integrity of the anastomosis is entirely dependent on the stent placed. Thus selection of an appropriate stent is paramount. Potential adverse events from transmural procedures are higher and include pneumoperitoneum, bleeding, cholangitis, stent dislocation, free perforation, bile leak, and bile peritonitis. Rarer complications include hemobilia, acute cholecystitis, duodenal double puncture, mediastinitis, and mortality.
Algorithm for Drainage and Choice of Procedure
The optimal algorithm for EUS-BD is still in evolution. It is unclear whether one type of EUS-BD procedure is preferred for a particular indication. The choice of procedure would need to balance on the technical and clinical success rates, stent patency, and risk profile of each procedure. Such an algorithm also should be individualized for each institution based on the technical expertise available. Our current approach is shown in Fig. 24.2 . The cause of failed ERCP should be examined. For benign causes, EUS-Rv ERCP should be attempted first. Transmural drainage should only be used with strong justification. For unresectable malignancies, transpapillary or antegrade procedures should be attempted first, depending on disease and anatomical considerations. When guidewire manipulation across the stricture is not possible, then transluminal approaches should be employed. In expert hands, both transpapillary and transluminal procedures appear to have similar efficacies and risks. Hence, transluminal procedures may be used more frequently. Park et al. proposed an algorithm based on accessibility of the papilla in patients with failed ERCP. A rendezvous procedure was performed first if the papilla was accessible. An antegrade procedure was then performed if the papilla was not accessible. Transluminal procedures were then performed if tumor invasion was present in the duodenum. Tyberg et al. categorized their patients based on cross-sectional imaging. If the intrahepatic ducts were dilated, an antegrade procedure was the first choice. If the ducts were not dilated, then a rendezvous procedure was performed. When these procedures failed, transluminal drainage was performed.
Endoscopic Ultrasonography-Guided Pancreatic Duct Drainage
Endoscopic retrograde pancreatography (ERP) is the conventional method for treating pancreatic ductal obstruction caused by strictures, stones, or congenital anomalies. The procedure may not be technically possible in about 3% to 10% of patients due to surgically altered anatomy, tight strictures, complete ductal obstructions, or disrupted ducts. Surgical interventions may be required for failed ERP. Recently, EUS-guided pancreatic duct drainage (EUS-PD) has been described as a rescue method for the management of patients in whom ERP is unsuccessful. However, EUS-PD seems to be one of the most technically demanding EUS-guided interventions, and it is associated with up to 43% risk of adverse events. Therefore it has only been performed by highly skilled endoscopists in tertiary centers with extensive experience in therapeutic ERP and EUS procedures.
Indications for Endoscopic Ultrasonography-Guided Pancreatic Duct Drainage
There are no guidelines as to when EUS-PD should be considered. Based on the previous literature and our experience, the potential indications for EUS-PD are summarized in Table 24.1 . Indications should be decided cautiously by taking into account the condition of the patient, the endoscopist’s experience, and the facilities available. Only experienced endoscopists trained in EUS and ERP should attempt this procedure.
EUS-PD can be classified into the EUS-guided rendezvous technique and EUS-guided transmural drainage ( Fig. 24.3 ). In the rendezvous technique (EUS-Rv pancreatic), EUS is used for pancreatic duct access and guidewire insertion to facilitate cannulation and stent placement in subsequent ERCP. When the papilla or anastomosis is inaccessible or unable to be traversed with the guidewire, EUS-guided transmural drainage can be performed. EUS-guided transmural PD can be divided into EUS-antegrade pancreatic duct drainage (EUS-AG PD) or EUS-guided pancreatico-gastrostomy (EUS-PG). Both procedures involve creation of a transmural fistula and transgastric placement of stents. In EUS-AG PD, antegrade placement of the stent traverses the papilla or surgical anastomosis. In EUS-PG, a pancreatico-gastrostomy is created with stent placement without traversing the papilla or anastomosis.
Outcomes of Endoscopic Ultrasonography-Guided Pancreatic Duct Drainage Procedures
EUS-PD is a challenging procedure. There have been few reports on the outcomes of EUS-Rv technique. The technical success rate of this procedure ranged from 25% to 100%. The outcomes of EUS-transmural PD in studies with more than 20 patients are shown in Table 24.7 . In recent systematic reviews of EUS-guided transmural PD, the overall technical success rate was 70% to 100%. The possible reasons for low technical success rates are (1) small diameter of pancreatic ducts; (2) hard sclerotic pancreatic parenchyma; (3) difficult guidewire manipulation in a tortuous main pancreatic duct (MPD); and (4) a lack of dedicated devices.
|Author||Year||Patients||Puncture Route||Type of Stent||Technical Success (%)||Clinical Success (%)||Early Adverse Events (%)||Late Adverse Events (%)||Reintervention Rate (%)|
|Tessier et al.||2007||36||TG/TB||PS||92||70||5%||None||55|
|Oh et al.||2016||25||TG/TB/TE||SEMS||100||100||5%||None||48|
|Tyberg et al.||2017||80||TG/TB||PS||81 a||81||20%||11%||N/A|
|Chen et al.||2017||40||N/A||PS||92.5 b||87.5||35%||None||N/A|