Endoscopic ultrasonography guided drainage: Summary of consortium meeting, May 21, 2012, San Diego, California

Abstract

Endoscopic retrograde cholangiopancreatography (ERCP) is the preferred procedure for biliary and pancreatic drainage. While ERCP is successful in about 95% of cases, a small subset of cases are unsuccessful due to altered anatomy, peri-ampullary pathology, or malignant obstruction. Endoscopic ultrasound-guided drainage is a promising technique for biliary, pancreatic and recently gallbladder decompression, which provides multiple advantages over percutaneous or surgical biliary drainage. Multiple retrospective and some prospective studies have shown endoscopic ultrasound-guided drainage to be safe and effective. Based on the currently reported literature, regardless of the approach, the cumulative success rate is 84%-93% with an overall complication rate of 16%-35%. endoscopic ultrasound-guided drainage seems a viable therapeutic modality for failed conventional drainage when performed by highly skilled advanced endoscopists at tertiary centers with expertise in both echo-endoscopy and therapeutic endoscopy

Core tip: This summary of the endoscopic ultrasound (EUS)-guided biliary drainage consortium held in 2012 focuses on technical improvements in both EUS-Guided biliary and pancreatic drainage techniques. This summary also provides a detailed overview of EUS-guided choledochoduodenostomy compared to percutaneous transhepatic cholangiography and surgical drainage. Other EUS Guided techniques such as endoscopic ultrasound-guided pancreatico-gastrostomy and endoscopic ultrasonography-guided cholecystoduodenostomy and cholecystogastrostomy have been discussed. Lastly, an extensive review of therapeutic endoscopic interventions in surgically altered anatomy has been provided as well.

INTRODUCTION

Endoscopic ultrasound (EUS) has evolved from a simple diagnostic procedure in the last two decades. This is in part due to the landmark study published by Wiersema[1], whom first used the EUS to guide a cholangiography to define the anatomy of the biliary tree. This shifting paradigm paper paved the ground for the subsequent evolution of the EUS as a powerful therapeutic tool. Giovannini et al[2] was the first article describing the use of EUS guiding biliary drainage, follow by Kahaleh et al[3] characterizing the “rendezvous” techniques.

Endoscopic transpapillary biliary drainage is the procedure of choice for biliary decompression in patients with unresectable pancreatic cancer associated to obstructive jaundice[3-6]. However, endoscopic retrograde cholangio-pancreatography (ERCP) failure can occur in 3%-10% of cases[3-5]. This failure can be due to operator inexperience, anatomic variation, tumor extension, prior surgery or incomplete drainage[3-9].

The indications for EUS-Guided biliary drainage include (1) failed conventional ERCP; (2) altered anatomy; (3) tumor preventing access into the biliary tree; and (4) contraindication to percutaneous access (i.e., ascites, etc.). The Consensus guidelines for the management of biliary obstruction using EUS guided biliary drainage were reviewed and updated to reflect practice patterns, techniques and new indications among the field.

TECHNIQUES APPROACHES IN EUS-GUIDED BILIARY DRAINAGE

Access

Biliary drainage guided by EUS can be performed at two locations, depending on the level of access to the biliary system.

Intrahepatic: The intrahepatic biliary system can be reached either transesophageal, transgastric or transjejunal (in altered anatomy), being the biliary segment III of the left hepatic lobe the most frequent and best visualized duct, especially when the EUS probe is placed at the stomach cardia and lesser curvature[9-11].

Extrahepatic: This technique can be performed having the needle accessing the common bile duct (CBD) directly, either using the transmural access from the antral part of the stomach or duodenum. In one hand, this approach provide not only a better visualization of the CBD, that some endoscopist considered advantageous at the time to select the route of access to the biliary system, but also due to the anatomical position of the CBD (located in the retroperitoneal space) might be safer in patients with ascites[10]. On the other hand, the major limitation for this technique is the difficulty inherent in the anterograde placement of the stent because of the angle of the needle entering the biliary duct.

Drainage

Rendezvous: Modality that allows advancing a guided wire through the papilla in antegrade fashion in which the bile duct is located and cannulated using the EUS rather than by retrograde cannulation with a duodenoscope. Technique especially used in those cases where biliary drainage is needed, ERCP failed and the duodenal anatomy allows the placement of the scope at the ampulla and the wire is identified traversing the papilla. Therefore, duodenal anatomy will determine the feasibility of the procedure, and that represent the main limitation[9].

Antegrade: In those cases where the luminal obstruction can not be overcome and the papilla is not visualized; but, the transpapillary wire access was obtained with EUS-guidance, then the alternative for biliary drainage is the antegrade placement of a biliary stent across the obstruction.

Transmural: Biliary drainage is obtained by creating a fistula in those cases where the wire cannot be positioned across the papilla due to either anomaly in the anatomy (biliary obstruction by a tumor) or technical complication (difficult position)[9].

ENDOSCOPIC ULTRASOUND-GUIDED BILIARY DRAINAGE

There is no consensus among the experts in the field of advance endoscopy regarding when to use EUS-BD, however, most of them can agree that failure in the use of conventional ERCP might be used as main explanation to justify the procedure, however, the indications for EUS-BD are not being established yet[12]. Therefore, EUS-BD has been rapidly being accepted as an alternative and reasonable option in those cases where bile duct cannulation cannot be achieved[11]. Either surgically altered anatomy (bariatric surgery or intestinal diversion for pancreatic cancer or other diseases) or obstruction of the gastrointestinal (GI) tract or bile duct (must often due to malignant causes) can be considered the principal causes for unsuccessful ERCP[13,14].

Most recent data reports that the accumulative success rate for extrahepatic EUS-BD (Table 1) is approximately 93% over the last 12 years[3,4,6,13-56], and in case of intrahepatic EUS-BD (Table 2) the cumulative success rate published is 84%[3,6,13,14,34,36,40,43,47,50-52,57]. However, data from two large multicenter retrospective trials failed to report advantages of any of these techniques[58,59].

Table 1.

Studies on extrahepatic endoscopic ultrasound-guided biliary drainage

Ref.	No./total sample	Method	Disease	Approach	Initial stent	Percent success rate	Complications
Giovannini et al[27] (2001)	1/1	Direct (1)	Malignant (1)	Duodenum	PS (1)	100	None
Burmester et al[6] (2003)	3/4	Direct (4)	Malignant (4)	Duodenum (2), stomach (1), jejunum (1)	PS (3)	75	Bile leak (1)
Mallery et al[28] (2004)	2/2	Rendez-vous (2)	Malignant (2)	Duodenum (2)	MS (2)	100	None
Lai and Freeman[29] (2005)	1/1	Rendez-vous (1)	Malignant (1)	Duodenum (1)	MS (1)	100	None
Püspök et al[13] (2005)	5/5	Direct (5)	Malignant (4), benign (1)	Duodenum (5)	PS (5)	80	Subacute phlegmonous cholecystitis (1)
Kahaleh et al[3] (2006)	10/23	Direct (2), rendez-vous (7)	Malignant (8), benign (2)	Duodenum (5), jejunum (5)	PS (4), MS (5)	90	Bile leak (1), pneumoperitoneum (2)
Yamao et al[20] (2006)	2/2	Direct (2)	Malignant (2)	Duodenum (2)	PS (2)	100	None
Ang et al[4] (2007)	2/2	Direct (2)	Malignant (2)	Duodenum (2)	PS (2)	100	None
Fujita et al[30] (2007)	1/1	Direct (1)	Malignant (1)	Duodenum (1)	PS (1)	100	None
Will et al[31] (2007)	8/8	Direct (8)	Malignant (7), benign (1)	Stomach (4), jejunum (3), esophagus (1)	PS (2), MS (5)	88	Slight pain (2), cholangitis (1)
Yamao et al[21] (2008)	3/3	Direct (3)	Malignant (3)	Duodenum (3)	PS (3)	100	Pneumoperitoneum (1)
Tarantino et al[19] (2008)	8/8	Direct (4), Rendez-vous (4)	Malignant (7), benign (1)	Duodenum (8)	PS (8)	100	None
Itoi et al[32] (2008)	4/4	direct (4)	Malignant (4)	Duodenum (4)	PS (3), NBD (1)	100	Focal peritonitis (1), bleeding (1)
Brauer et al[33] (2009)	12/12	Direct (4), Rendez-vous (7)	Malignant (8), benign (4)	NA	PS (5), SEMS (5)	92	Pneumoperitoneum (1), respiratory failure (1)
Horaguchi et al[14] (2009)	9/16	NA	Malignant (9)	Duodenum (8), stomach (1)	PS (14), plastic PT (1), NBT (1)	100	Peritonitis (1)
Hanada et al[17] (2009)	4/4	Direct (4)	Malignant (4)	Duodenum (4)	PS (4)	100	None
Maranki et al[34] (2009)	14/49	Direct (6), Rendez-vous (8)	Malignant (9), benign (5)	NA	NA	86	Biliary peritonitis (1), abdominal pain (1), pneumoperitoneum (1)
Kim et al[35] (2010)	15/15	Rendez-vous (15)	Malignant (10), benign (5)	Duodenum (15)	PS (4), MS (8)	80	Pancreatitis (1)
Nguyen-Tang et al[36] (2010)	1/5	Rendez-vous (1)	Malignant (1)	NA	MS (1)	100	None
Iwamuro et al[37] (2010)	7/7	Direct (7)	Malignant (7)	Duodenum (5), stomach (2)	PS (7)	100	Bile peritonitis (2)
Artifon et al[38] (2010)	3/3	Direct (3)	Malignant (3)	Duodenum (3)	MS (3)	100	None
Belletrutti et al[39] (2010)	1/1	Direct (1)	Malignant (1)	Duodenum (1)	MS (1)	100	None
Park do et al[40] (2011)	31/57	Direct (31)	Malignant (51), benign (6)	Duodenum (31)	PS (6), MS (25)	87	Pneumoperitoneum (6), mild bleeding (2)
Fabbri et al[41] (2011)	16/16	Direct (13), Rendez-vous (3)	Malignant (16)	Duodenum (15), stomach (1)	PS (4), MS (8)	80	Pancreatitis (1)
Hara et al[42] (2011)	18/18	Direct (18)	Malignant (18	NA	PS (17)	94	Peritonitis (2), bleeding (1)
Ramírez-Luna et al[43] (2011)	9/11	Direct (9)	Malignant (9)	Duodenum (9)	plastic DPT (9)	89	Biloma (1)
Siddiqui et al[44] (2011)	8/8	Direct (8)	Malignant (8)	Duodenum (80	MS (8)	100	Stent migration (1), duodenal perforation (1)
Komaki et al[45] (2011)	15/15	Direct (14), Rendez-vous (1)	Malignant (15)	Duodenum (15)	PS (15)	100	None
Prachayakul et al[46] (2011)	1/1	Direct (1)	Malignant (1)	Duodenum (1)	PS (1)	100	None
Artifon et al[22] (2012)	13/13	Direct (13)	Malignant (13)	Duodenum (13)	MS (13)	100	Bile leak (1), bleeding (1)
Attasaranya et al[47] (2012)	10/31	Direct (9), Antegrade (1)	Malignant (23), benign (8)	Duodenum	NA	60	4/10 (40%)
Katanuma et al[48] (2012)	1/1	Direct (1)	Benign (1)	Duodenum (1)	PS (1)	100	None
Kawakubo et al[49] (2012)	2/2	Direct (2)	Malignant (2)	Duodenum (2)	PS (2)	100	None
Khashab et al[50] (2012)	7/9	Direct (2), antegrade (2), Rendez-vous (3)	Malignant (7)	Duodenum (6), gastric (1)	MS (7)	100	Pancreatitis (1), cholecystitis (1), abdominal pain (1)
Kim et al[51] (2012)	9/13	Direct (9)	Malignant (9)	Duodenum (9)	MS (9)	100	Pneumoperitoneum (2), migration (2), peritonitis (1)
Song et al[52] (2012)	15/15	Direct (15)	Malignant (15)	Duodenum (15)	MS (15)	100	Pneumoperitoneum (2), cholangitis (1)
Dhir et al[53] (2012)	58/58	Rendez-vous (58)	Malignant (43), benign (15)	Duodenum (58)	NA	98	Bile leak (2)
Hara et al[54] (2013)	18/18	Direct (18)	Malignant (18)	Duodenum (18)	MS (18)	94	Bile peritonitis (2)
Park do et al [55] (2013)	16/45	Direct (2), Rendez-Vous (14)	Malignant (39), benign (6)	Duodenum (16)	MS (16)	88	Pancreatitis (1), bile peritonitis (1)
Itoi et al [56] (2013)	1/1	Direct (1)	Malignant (1)	Stomach (1)	MS (1)	100	None
Total	365					338/365 (93%)	58/365 (16%)

Table 2.

Studies on intrahepatic endoscopic ultrasound-guided biliary drainage

Ref.	No./total sample	Method	Disease	Approach	Initial stent	Percent success rate	Complications
Burmester et al[6] (2003)	1/4	Direct (1)	Malignant (1)	Stomach (1)	PS (1)	100	Bile leak (1)
Püspök et al[13] (2005)	1/1	Direct (1)	Malignant (1)	jejunum (1)	PS (1), MS (1)	100	None
Kahaleh et al[3] (2006)	13/23	Direct (1), Rendez-vous (12)	Malignant (9), benign (4)	Stomach (13)	PS (6), MS (6)	92	Minor bleeding (1)
Bories et al[16] (2007)	11/11	Direct (9), Antegrade (2)	Malignant (3), benign (8)	Stomach (3), Duodenum (3), stenosis (5)	PS (7), MS (3)	91	Transient ileus (1), biloma (1), cholangitis (1)
Horaguchi et al[14] (2009)	7/16	NA	Malignant (7)	Stomach (5), esophagus (2)	PS (2), MS (5)	100	None
Maranki et al[34] (2009)	35/49	Direct (9), Antegrade (24)	Malignant (26), benign (9)	NA	NA	83	Bleeding (1), pneumoperitoneum (3), aspiration pneumonia (1)
Nguyen-Tang et al[36] (2010)	4/5	Rendez-vous (4)	Malignant (3), benign (1)	Duodenum (1), Stomach (3)	MS (5)	100	None
Park do et al[40] (2011)	31/57	Direct (31)	Malignant (51), benign (6)	Duodenum (31)	PS (6) MS (25)	87	Pneumoperitoneum (1), bile peritonitis (2)
Ramírez-Luna et al[43] (2011)	2/11	Direct (2)	Malignant (2)	Stomach (2)	PS (2)	100	Stent migration (1)
Attasaranya et al[47] (2012)	16/31	Direct (16)	Malignant (23), benign (8)	NA	NA	81	6/16 (38%)
Khashab et al[50] (2012)	2/9	Antegrade (1), rendez-vous (1)	Malignant (2)	Stomach	MS (2)	100	Nausea (1)
Kim et al[51] (2012)	4/13	Direct (4)	Malignant (4)	Stomach (4)	MS (3)	75	Peritonitis (1), stent migration (1)
Park do et al[55] (2013)	29/45	Direct (9), antegrade (14), rendez-vous (5)	Malignant (39), benign (6)	Stomach (29)	NA	66	Biloma (1)
Iwashita et al[57] (2013)	6/6	Direct (1)	Malignant (1), benign (5)	NA	MS (1)	100	Pancreatitis (1), abdominal pain (1)
Total	162					136/162 (84%)	26/162 (16%)

EUS-guided rendezvous

This procedure is performed inserting a needle under EUS and doppler guidance into either left intrahepatic duct or common bile duct. Here, the FNA needle will permit also the advance of guidewire distally into the duodenum. Upon insertion of the needle into the duct, suspected by EUS imaging, the aspiration of the bile will confirm the intraductal placement of the needle, which is follow by injection of contrast drawing the biliary tree (cholangiogram). Next, the insertion of the guidewire through the FNA needle toward the duodenum and perform the conversion to a conventional ERCP in a retograde fashion. If by any reason the transpapillary passage is not successfully accomplish, then the FNA needle need to be replaced by the sphincterotome or dilating bougie, allowing the manipulation of the wire freely and safely to facilitate the passage through the ampulla. At this point, the EUS scope is removed leaving the wire in the duodenum and the duodenoscope is advanced to the ampulla where the guidewire can be grasped using a snare or forceps and pulled it back through the working channel of the duodenoscope and permit the subsequent over-the-wire cannulation or alternately use the duodenoscope to cannulate the common bile duct next to the prior placed guidewire. Now the procedure can be completed by conventional endoscopic retrograde cholangiography with stent placement in retrograde manner[11].

EUS-guided choledochoduodenostomy vs other techniques (percutaneous transhepatic cholangiography and surgical drainage)

Burmester et al[6] described a one-step method using a new device consisted of a 19 gauge fistulotome with a 0.25-inch guidewire, a pusher tube and an 8.5F plastic stent fixed with a 3.0 nylon-suture. This method of direct puncture of the extrahepatic or an intrahepatic duct could reduce the risk of guidewire dislocation during the instrument change, what must be made with the two-step method[17]. However, more studies with this device are needed.

EUS-guided biliary drainage has many advantages over PTBD[3,4,16,17]. The proximity of the transducer to the bile duct during EUS is the major advantage[3,19]. Even in patients who have undergone total gastrectomy or partial gastrectomy with a Billroth II reconstruction, EUS can reveal the etiology of extrahepatic cholestasis, situations in that ERCP may not be possible[1,3,4,15,16]. Other advantages include puncture of the biliary tree with color-Doppler information to avoid vascular injury, the lack of ascites in the interventional field and the lack of an external tube, improving the quality of life of the patients[4,20].

Choledochoduodenostomy can prevent clogging and tumor ingrowth and/or overgrowth, because it creates a fistula far from the obstructing tumor[20,21]. Many studies described this procedure with high success rates (more than 90%) and low rate of procedure-related complications (around 19%)[18]. The main risk of EUS-guided biliary drainage is bile leakage, especially if stent insertion is unsuccessful[4]. Burmester et al[6] reported the failure of stent placement in 1 of their 4 patients, causing bile peritonitis[17]. They also reported that only local peritonitis developed, which did not contribute to the death of the patient. Some investigators recommend the transhepatic approach to decrease the risk of biliary peritonitis in case of stent failure[17]. Other complications include pneumoperitoneum and minor bleeding[4,5,1].

EUS-guided choledochoduodenostomy for malignant biliary obstruction has been shown to be an effective alternative to PTBD or surgery when ERCP fails. Artifon et al[22] compared EUSCD and PTC in 25 patients with distal biliary malignant obstruction. The 2 groups were similar before intervention in terms of quality of life [EUS-CD (58.3) vs PTBD (57.8), P = 0.78], total bilirubin (16.4 vs 17.2, P = 0.7), alkaline phosphatase (539 vs 518, P = 0.7), and gamma-glutamyl transferase (554.3 vs 743.5, P = 0.56). All procedures were technically and clinically successful in both groups. The study concluded that EUS-CD can be an effective and safe alternative to PTBD with similar success, complication rate, cost, and quality of life.

Another Study conducted by Artifon et al[23] showed results on comparative trial between EUS CD and surgery to patients with biliary distal cancer. There was no significant difference in the technical and clinical outcomes in the two groups. Cost analysis demonstrated a significantly increased cost per patient in the surgical group (P = 0.039). Complications were significantly higher in the surgical group (P = 0.041). There was one case of self-limited bleeding on group EUSCD and one case each of wound abscess, abdominal abscess, internal fistula and pneumonia on group Surgery. All patients in surgical group were managed conservatively.

ENDOSCOPIC ULTRASOUND-GUIDED PANCREATICO-GASTROSTOMY

Main indications are stenosis of pancreatico-jejunal or pancreatico-gastric anstomosis after Whipple resection, which induce recurrent acute pancreatitis, main pancreatic duct stenosis due to chronic pancreatitis, post-acute pancreatitis or post pancreatic trauma after failure of ERCP. EUS guided pancreatico-gastro or bulbostomy offers an alternative to surgery (Figure 1).

Figure 1.

Pancreatico-gastrostomy/stenosis of a wirsung-gastro anastomosis after whipple resection for benign cystic lesion of the head of the pancreas.

By using a linear interventional echoendoscope, the dilated main pancreatic duct (MPD) was well visualized. Endoscopic pancreatico-gastrostomy (EPG) was then performed under combined fluoroscopic and ultrasound guidance, with the tip of the echoendoscope positioned such that the inflated balloon was in the duodenal bulb while the accessory channel remained in the antrum. A needle was inserted transgastrically into the proximal pancreatic duct and contrast medium was injected. Opacification demonstrated a pancreaticogram the needle was exchanged over a guidewire, which was then used to enlarge the channel between the stomach and MPD. The sheath was introduced by using cutting current. After exchange over a guidewire (rigid 0.035 inch diameter), a 7F, 8-cm-long pancreaticogastric stent was positioned. This stent will be exchanged for two 7F or one 8.5F stents one month after the first procedure.

The results of the fours series[3,24-26] of patients published are much too preliminary in nature to recommend wider use of EPG, which in any case should be restricted to tertiary centers specializing in biliopancreatic therapy with a pain relief in 70% of cases. But the complication rate is still high around 15% including bleeding, pancreatic collection and perforation. The largest series was published by Tessier et al[24] on 36 patients. Indications were chronic pancreatitis, with complete obstruction (secondary to a tight stenosis, a stone, or MPD rupture); inaccessible papilla or impossible cannulation (n = 20); anastomotic stenosis after a Whipple procedure (n = 12); complete MPD rupture after acute pancreatitis (AP); or trauma (n = 4). EPG or EPB was unsuccessful in 3 patients; 1 was lost to follow-up. Major complications occurred in 2 patients and included 1 hematoma and 1 severe AP. The median follow-up was 14.5 mo (range: 4-55 mo). Pain relief was complete or partial in 25 patients (69%, intention to treat). Eight patients treated had no improvement of their symptoms (4 were subsequently diagnosed with cancer). Stent dysfunction occurred in 20 patients (55%) and required a total of 29 repeat endoscopies. The last study published by Ergun et al[26] reported the long-term follow-up (37 mo) of 18 patients who underwent a EUS guided pancreatico-gastrostomy. Stent occlusion occurred in 50%, pain relief was always present in 70% and the mean pain score decreased dramatically from 7.5 to 1.6.

It’s very difficult to find today the place of EGD in our experience the best indication is anastomotic stenosis after Whipple procedure for benign pancreatic lesions (cystadenoma, IPMN, NET). EGD offers an alternative to surgery and the best results in the 3 series published (Table 3) were showed in this indication. In another hand, surgery should be considered as the elective treatment of CP after failure of the endoscopic route.

Table 3.

Studies on endoscopic ultrasound-guided pancreatico-gastrostomy

Ref.	NB PTS	Success	Complication	Follow-up
Tessier et al[24] (2007)	36	70%	11%	16.5 mo
Kahaleh et al[3] (2006)	13	92%	16%	14 mo
Barkay et al[25] (2010)	21	48%	2%	13 mo
Ergun et al[26] (2011)	20	90%	10%	37 mo

ENDOSCOPIC ULTRASONOGRAPHY-GUIDED CHOLECYSTODUODENOSTOMY AND CHOLECYSTOGASTROSTOMY

Revised Tokyo Guideline (TG13) proposes that early or emergency cholecystectomy should be conducted as the gold standard of treatment for acute cholecystitis[60-63]. In general, cholecystectomy is relatively safe. However, the mortality rate of cholecystectomy in patients at high risk due to comorbid conditions is not lower than in non-critical ill patients[64,65]. Therefore, several literatures advocates that high-risk patients should be tentatively treated by decompress the gallbladder, e.g., percutaneous transhepatic gallbladder drainage (PTGBD), percutaneous transhepatic gallbladder aspiration, endoscopic transpapillary gallbladder drainage[66-69]. Recently, as a novel translumenal access, endoscopic ultrasonography-guided gallbladder drainage (EUS-GBD) has been performed.

Technique of EUS-guided gallbladder drainage

The gallbladder is visualized from the duodenal bulb or the antrum of the stomach using a curved linear array echoendoscope in a long scope position (pushing scope position)[69].

After depiction of neck and body of the gallbladder, a 19-gauge needle is inserted transduodenally or transgastrically into the gallbladder under EUS visualization. After the stylet is removed, bile is aspirated approximately 3 cc and same 3 cc contrast medium is then injected to confirm the tip of a needle is in the gallbladder. Then, a 0.025-inch or 0.035-inch guidewire was inserted through the 19-gauge needle and looped in the gallbladder. A biliary catheter for dilation up to 8-Fr or 4-mm to 6-mm papillary balloon dilator if necessary, are used for dilation of the cholecystoentero fistula. Finally, a 5-8.5-Fr pigtail type naso-gallbladder drainage catheter, double pigtail type plastic stent, or fully-covered self-expandable metal stent are inserted through the gastric or duodenal fistula into the gallbladder.

Ten reports describe the outcome of EUS-guided gallbladder drainage (Table 4)[70-81]. In total, the technical success and clinical success rate were 98.7% (74/74) and 100% (74/74), respectively. As a first puncture, a 19-gauge needle was mainly used. A 5- to 8.5-F plastic stents or naso-gallbladder catheters were used for drainage. In 2 institutions, dedicated fully-covered metal stents, which had double flanges were used.

Table 4.

Summary of published data on endoscopic ultrasound-guided gallbladder drainage

Ref.	No. of cases	Device for puncture	Approach route	Technical success	Clinical success	Stent	Complication (No. of cases)
Baron et al[70] (2007)	1	19 G FN	TD	100%	100%	7F PS	None
Kwan et al[71] (2007)	3	19 G FN/FT/CT	TD	100%	100%	8.5F NBD	Bile leakage (1)
Lee et al[72] (2007)	9	19 G FN	TD	100%	100%	5F NBD	Pneumoperitoneum (1)1
Kamata et al[74] (2009)	1	19 G FN	TG	100%	100%	7F PS	None
Takasawa et al[73] (2009)	1	NK	TG	100%	100%	7F PS	None
Song et al[75] (2010)	8	19 G FN +/- NK	1TG/7TD	100%	100%	7F PS	Bile Leakage (1), pneumoperitoneum (1), stent migration (1)
Súbtil et al[81] (2010)	4	CT/NK	TG	100%	100%	8.5F PS	None
Jang et al[77] (2011)	15	19 G FN/NK	10TG/5TD	100%	100%	New SEMS	Pneumoperitoneum (2)
Jang et al[79] (2012)	30	19 G FN/NK	TG/TD	97%	100%	5F NBD	Pneumoperitoneum (2)
Itoi et al[78] (2012)	5	19 G FN/1 CT	1TG/4TD	100%	100%	New SEMS	Mild bleeding (1)
de la Serna-Higuera et al[80] (2013)	13	CT	12TG/1TD	84.6%	100%	New SEMS	Hematochezia (1/11), pain (1/11)

¹Minor bile leakage without serious bile peritonitis. G: Gauge; NK: Needle knife; FT: Fistulotome; CT: Cystotome; PS: Plastic stent; NBD: Naso-Gallbladder Drain; TD: TransDuodenal; Approach TG: TransGastric approach; Partially covered SEMS: Partially cover self expandable metallic stent.

As an adverse events, 6 of self-limited pneumoperitoneum, 2 of bile leakage, and 1 distal stent migration without bile leakage 3 wk after stent insertion. Interestingly, Jang et al[79] conducted randomized control study, PTGBD vs EUS-guided naso-gallbladder drainage (EUS-NGBD). EUS-NGBD and PTGBD showed similar technical [97% (29/30) vs 97% (28/29); 95%] and clinical [100% (29/29) vs 96% (27/28)] success rates, and similar rates of complications [7% (2/30) vs 3% (1/29)]. The median post-procedure pain score was significantly lower after EUS- NGBD than after PTGBD (1 vs 5, P < 0.001). On the basis of the results, they suggested that EUS-NGBD is comparable with PTGBD in terms of the technical feasibility and efficacy; there were no statistical differences in the safety. EUS-GBD is a good alternative for high-risk patients with acute cholecystitis who cannot undergo an emergency cholecystectomy.

EUS-guided gallbladder drainage has been developed as an alternative drainage method for acute cholecystitis. However, there are several limitations of this procedure. One of the big issues is that basically there is no adhesion and relatively distance between the GI tract including gastric and duodenal wall and the gallbladder wall. In addition, tubular biliary plastic and metal stents for biliary decompression have significant shortcomings when used for transenteric drainage. Plastic stents have the disadvantage of a small lumen diameter, which can limit drainage and may necessitate reintervention. Currently available SEMSs have a larger lumen diameter but may show inward and outward migration. Furthermore, abutment of the end of a tubular SEMS against the lumen wall may cause tissue injury and bleeding. Although serious adverse events have never been reported, we should consider the possibility of such unexpected events.

Recently, Itoi et al[78] de la Serna-Higuera et al[80] and Jang et al[79] reported a newly designed fully-covered metallic stent with anchor for EUS-guided gallbladder drainage. These dedicated devices may suggest that this procedure is becoming safer and more reliable. That causes reducing the risk of serious complications like bile leakage and stent migration.

SURGICALLY ALTERED ANATOMY

Surgery of the upper GI tract involving change in the configuration of the antro-duodenal C-loop is commonly referred to as postoperative or surgically altered anatomy (SAA). Obesity surgery, peptic ulcer disease, iatrogenic bile-duct injury and chronic pancreatitis are common benign causes of SAA. Gastric, biliary and pancreatic malignancies also lead to radical resections with curative intent or to palliative bypass procedures resulting in SAA. Interventions performed for both benign and malignant disease include Billroth II, variations on the Whipple operation, and Roux-en-Y reconstructions, either with intact papillae[82] or with bilio-enteric anastomoses[83]. Most of these interventions result in long afferent limbs of small bowel (jejunum) impairing endoscopic access to the papilla.

Indications for biliary drainage in SAA patients

More insidious biliary obstructive symptoms caused by recurrent malignancy are seen in afferent limb syndrome, a condition difficult to diagnose and challenging to treat[84]. Obesity and weight loss are associated with common bile duct stones. Gallstones are almost always present in patients undergoing cholecystectomy, which is the major cause of bile-duct injury. Chronic reflux of GI contents across widely patent bilio-enteric anastomoses causes sump syndrome and stone formation. Leakage or anastomotic strictures may develop acutely after bilio-digestive bypass surgery. Anastomotic strictures more commonly cause chronic cholestasis and/or recurrent cholangitis over the long-term, with a reported incidence up to 30% after a mean follow-up of six years[85].

Dominant ERCP strategies in SAA

Access to the papilla and ease of en-face view are nearly normal in some cases with past SAA for benign conditions, such as Billroth I for peptic ulcer or choledochoduodenostomy for common bile duct stones. Nowadays, experienced operators report similarly high success rates for duodenoscope-based ERCP in patients with Billroth II. Whipple and Roux-en-Y reconstructions, particularly Roux-en-Y gastric by-pass (RYGB) for obesity, have ultra-long afferent limbs. Special endoscopes are usually required to access the proximal afferent limb for ERCP in these patients. Balloon or spiral enteroscopes are currently used at expert centers[86]. Enteroscopy-based ERCP is nonetheless a labor-intensive procedure with limited overall success rates of 60%. Seventy-five percent of failures occur because the papilla or bilio-enteric anastomosis cannot be identified, cannulated or drained, even despite successful limb intubation[86,87]. These figures underscore the limitations to ERCP using long, narrow channel, forward viewing enteroscopes.

An alternative approach to enteroscopy-based ERCP in long limb SAA patients is percutaneous transenteric access under open[6] or laparoscopic[88] surgical assistance. Typically, laparoscopic surgeons access the distal gastric remnant of RYGB, allowing intraoperative transgastric insertion of a duodenoscope and antegrade ERCP[87-89]. Bariatric procedures involving distal gastrectomy are less readily amenable to laparoscopically-assisted ERCP, often requiring open surgery for duodenoscope passage[87]. The complex logistics of intraoperative ERCP help understand its limited dissemination. It should be noted nonetheless that higher success rates have been reported with laparoscopy-assisted transgastric ERCP compared to enteroscopy ERCP in RYGB patients, especially when Roux limbs are longer than 150 cm from the ligament of Treitz to the jejuno-jejunal anastomosis[90].

Non-standard approaches to ERCP in SAA

Prior indwelling transenteric feeding tubes may allow percutaneous ERCP through mature tracks in the occasional SAA patient. A standard duodenoscope was passed percutaneously into the jejunum after jejunostomy track dilation and ERCP with sphincterotomy was performed uneventfully in a patient with Roux-en-Y gastrectomy[91]. However, a two-step approach is usually required in RYGB. Gastrostomy to the distal gastric remnant is first performed for the purpose of providing a conduit for duodenoscope passage, so that elective ERCP can be carried out through the mature gastrostomy track 2-4 wk later. In this two-step approach, percutaneous gastrostomy to the remnant stomach can be performed by surgery[92-94], interventional radiology[95] or endoscopy[96]. In keeping with data from the intraoperative transgastric approach, ERCP success rates using this two-step percutaneous transgastric approach are significantly higher than success rates for enteroscopy-based ERCP[94]. Complications related to the gastrostomy are however not negligible[93,94]. Successful therapeutic ERCP through gastrostomy to the remnant stomach placed by interventional radiology has been reported in a single RYGB case[95]. However, radiographically guided percutaneous access to the non-distended remnant stomach entails risk and difficulty. EUS-guided puncture and distension of the remnant stomach from the proximal gastric pouch by injecting contrast and air appears to facilitate percutaneous radiographic gastrostomy[97]. Double-balloon enteroscopy retrograde access to the remnant stomach for PEG placement prior to transgastric antegrade ERCP appears a reasonable option, given the high failure rate of enteroscopy-based ERCP even after successful limb intubation. Retrograde enteroscopy PEG placement has been reported as the initially chosen approach for ERCP in RYGB[96], and it might become a convenient same-session salvage procedure in cases of failed enteroscopy ERCP despite successful limb intubation.

Antegrade endoscopic access to the gastric remnant through spontaneous gastro-gastric communications across the staple line has successfully used for ERCP in two RYGB patients. The defect was large enough to allow passage of a duodenoscope for sphincterotomy and stone removal in one case[98]. Another patient, however, required balloon dilation. A covered self-expandable esophageal stent was placed across the staple line to provide a conduit for iterative ERCP. Once final resolution of a benign biliary stricture was achieved, the transgastric covered stent was removed[99]. The concept of temporary covered metal stents serving as conduits for endoscope passage and therapeutic ERCP has also been applied to the percutaneous[96,100,101] or transluminal EUS-guided routes[102].

Afferent loop syndrome resulting in chronic biliary obstruction caused by recurrent malignancy is an increasingly frequent clinical problem. Palliation often involves permanent percutaneous drains, which adversely impact quality of life[84]. A novel EUS-based approach was used at two different institutions in three patients. The distended afferent loop was punctured from the distal antrum or duodenum, and a double plastic pigtail placed transmurally. Cholangitis or cholestasis resolved in all three case[103,104].

EUS-guided access and drainage routes in SAA

EUS-guided biliary drainage (EUSBD) is carried out by any of three possible routes, transmural, transpapillary antegrade or transpapillary retrograde (rendezvous) encompassed under a procedure hybrid between EUS and ERCP known as endosonography-guided cholangiopancreatography (ESCP)[105]. Patients with SAA represent around a fourth of cases in both current[55,58,106] and early[2,6,13] ESCP series inclusive of both intrahepatic and extrahepatic EUS-guided access (Table 5). Patients with SAA are not present however in ESCP series reporting predominantly on choledocho-duodenostomy or extrahepatic rendezvous, which represent two thirds of the total ESCP cases reported[105]. Direct drainage routes (transmural and antegrade transpapillary) are typically chosen for EUSBD in patients with SAA instead of rendezvous. The ability to endoscopically reach the papilla is a requisite for rendezvous, and in most patients with SAA the reason for ERCP failure dictating the need for ESCP is precisely failed access to the papilla. EUS rendezvous has nonetheless been attempted in patients with past Billroth II from an extrahepatic[107] or an intrahepatic[108] approach to extract CBD stones or to relieve malignant obstruction.

Table 5.

Outcomes of biliary drainage in endosonographic cholangiopancreatography series including surgically altered anatomy patients n (%)

Ref.	n	Patients with SAA	Overall technical success	Type of ESCP drainage				Complications
				Transpapillary		Transmural
				RV	AG	HG	CD
Will et al[31] (2007)	8	7 (87)	7 (87)	1	-	6	-	2 (25)
Bories et al[16] (2007)	11	4 (36)	10 (91)	-	-	10	-	4 (36)
Horaguchi et al[14] (2009)	16	4 (25)	16 (100)	-	-	8	8	2 (12)
Maranki et al[34] (2009)	49	7 (14)	41 (84)	201	141	3	4	8 (16)
Shah et al[106] (2012)	68	19 (28)	58 (85)	392	102	8	1	6 (9)
Iwashita et al[118] (2012)	40	9 (40)	29 (73)	29	-	-	-	5 (13)
Park et al[55] (2013)	45	14 (31)	41 (95)	16	9	14	2	5 (11)
Total	237	64 (27)	202 (85)	105	33	49	15	32 (13)
				138 (68)		64 (32)

¹Not specified in the original report, data provided by the author;

²Not specified in the original report. Data estimated from prior detailed series[48]. SAA: Surgically altered anatomy; ESCP: Endosonographic cholangiopancreatography; RV: Rendezvous; AG: Antegrade; HG: Hepatico-gastrostomy; CD: Choledochoduodenostomy.

The dominant EUS-guided access route in SAA patients is intrahepatic into the left hepatic duct branches from the cardia region: either transgastric from the proximal stomach, transesophageal from the abdominal esophagus, or transjejunal in patients with esophago-jejunostomy[31]. Even if left intrahepatic duct branches are less obvious target for EUS-guided puncture than the CBD, the rationale for it in SAA patients is strong. SAA frequently involves distal gastrectomy, as in Roux-en-Y gastrectomy for gastric cancer, or an excluded distal stomach, as in RYGB for obesity. As the CBD is typically imaged under EUS from the distal antrum or the duodenal bulb, extrahepatic EUS-guided access is usually not possible in SAA. An exception to this rule would be the occasional patient with distal biliary obstruction in whom the common hepatic duct can be imaged and accessed under EUS from the mid stomach. This frequently overlooked transgastric access site into the proximal extrahepatic bile duct may be useful in those SAA patients without dilated intrahepatics. Kahaleh et al[107] reported this option in 2 out of their first 5 cases. Another potentially useful secondary EUS-guided access route in highly selected SAA patients with intact stomach and proximal biliary obstruction but no left intrahepatic dilation, would be transduodenal from the bulb into the right hepatic duct or its branches. This recently reported access site was chosen in 6 cases with selective right intrahepatic dilation not amenable to standard left intrahepatic or extrahepatic access. Three of these 6 patients had Roux-en-Y hepaticojejunostomy. Hepaticoduodenostomy and antegrade transanastomotic intervention (stenting and balloon dilation) were used to carry out drainage[109]. EUSBD in patients with SAA usually involves transmural or antegrade transpapillary (or transanastomotic) stenting after left intrahepatic duct puncture. Exceptions to these general rules may nonetheless allow salvage of specific patient subsets, such as secondary access sites for those without left intrahepatic dilation[109] or retrograde rendezvous drainage for those requiring cross-over after failed antegrade or transmural ESCP approaches[55,106]. EUSBD has been reported to date in around a hundred patients with SAA. Only two small series focus specifically on EUSBD in SAA[57,110], patient specifics can be traced to procedural approach and outcomes in just 47 cases, 22 malignant (Table 6) and 25 benign (Table 7).

Table 6.

Patients with malignant biliary obstruction and surgically altered anatomy drained by endosonographic cholangiopancreatography

Ref.	No. of patients1	Type of SAA				Etiology			Type of ESCP drainage2				Success/complex4
									Transpapillary		Transmural
		RYG	RYHJ	PD	Other	Gast	Panc	Other	RV	AG	HG	Other3	Sn	Cn
Burmester et al[6] (2003)	2/4	1	-	-	1	1	1	-	-	-	1 (0)	1 (0)	2/2	0
Giovannini et al[2] (2003)	1/1	1	-	-	-	1	-	-	-	-	1 (1)5	-	1/1	0
Püspök et al[13] (2005)	1/6	1	-	-	-	1	-	-	-	1 (1)	-	1 (0)6	1/1	0
Kahaleh et al[3] (2006)	2/23	-	-	-	2	-	1	1	2 (1)	-	-	-	2/2	0
Will et al[31] (2007)	6/8	4	1	1	-	4	-	2	-	-	3 (3)	3 (1)	5/6	2
Bories et al[16] (2007)	1/11	-	-	1	-	-	1	-	-	-	1 (0)	-	1/1	0
Horaguchi et al[14] (2009)	4/16	-	2	-	2	1	1	2	-	-	2 (2)5	2 (1)5	4/4	1
Chopin-Laly et al[115] (2009)	1/1	-	-	-	1	-	1	-	-	-	1 (1)	-	1/1	0
Nguyen-Tang et al[36] (2010)	1/5	-	1	-	-	-	-	1	-	1 (1)	-	-	1/1	0
Ma et al[117] (2011)	1/1	-	-	1	-	-	-	1	-	-	1 (0)	-	1/1	0
Henry et al[116] (2011)	1/1	-	-	-	1	-	-	1	-	-	-	1 (0)7	1/1	1
Iwashita et al[118] (2012)	1/7	1	-	-	-	1	-	-	-	1 (1)	-	-	1/1	1
Total	22	8	4	3	7		5	8	2 (1)	3 (3)	10 (7)	7 (1)
		22				9 (40%)	13 (60%)		5 (23%)		17 (77%)		21 (95%)	5 (23%)

¹Number of surgically altered anatomy (SAA) patients with malignant biliary obstruction per total number of patients in the series;

²Number of patients with attempted stent placement (ITT), followed by number of metal stents in parentheses;

³Transmural variants of HG (hepatico-esophagostomy/jejunostomy) following intrahepatic access, except when stated otherwise;

⁴Number of successful cases or complications per number of SAA patients with malignant biliary obstruction;

⁵Initial plastic stent replaced at follow-up. Only final stent tallied in the total count;

⁶Additional hepato-esophageal plastic stent combined with AG metal stent in recurrent gastric cancer after RYG. Not tallied in the total count;

⁷Choledochoduodenostomy after palliative gastrojejunostomy. SAA: Surgically altered anatomy; ESCP: Endosonographic cholangiopancreatography; RYG: Roux-en-Y gastrectomy; RYHJ: Roux-en-Y hepatico-jejunostomy; PD: Pancreatico-duodenectomy; Gast: Gastric; Panc: Pancreatic; RV: Rendezvous; AG: Antegrade; HG: Hepatico-gastrostomy; ITT: Intention-to-treat; Sn: Successful cases; Cn: Complications.

Table 7.

Patients with benign biliary obstruction and surgically altered anatomy drained by endosonographic cholangiopancreatography

Ref.	No. of patients1	Type of SAA				Etiology		Type of ESCP drainage					Success/complex5
								Transpapillary2			Transmural3
		RYG	RYGB	RYHJ	Other	Stone	Strx	RV	AGS	AGB	HG	Other4	Sn	Cn
Püspök et al[13] (2005)	1/6	-	-	-	1	1	-	-	-	-	-	1 (0)6	1/1	0
Kahaleh et al[3] (2006)	2/23	-	-	1	1	1	1	2	-	-	-	-	1/2	1
Will et al[31] (2007)	1/8	-	-	1	-	-	1	17	-	-	-	1 (0)	1/1	0
Bories et al[16] (2007)	3/11	-	-	-	3	-	3	17	-	-	3 (1)	-	2/3	1
Weilert et al[110] (2011)	6/6	-	6	-	-	6	-	2	-	4	-	-	6/6	1
Artifon et al[112] (2011)	1/1	-	-	1	-	-	1	-	1	19	-	-	1/1	0
Park et al[113] (2012)	1/1	-	-	1	-	-	1	-	-	1	-	-	1/1	0
Bapaye et al[114] (2012)	1/1	-	-	1	-	-	1	-	-	1	-	-	1/1	0
Iwashita et al[57] (2013)	6/7	4	-	-	2	5	1	-	-	59	-	-	6/6	1
Park et al[109] (2013)	3/6	-	-	3	-	-	3	-	1	38	-	-	2/3	0
Total	25	4	6	8	7			6	2	13	3 (1)	2 (0)
		25				13 (52%)	12 (48%)	21 (84%)			5 (20%)		22 (88%)	4 (16%)

¹Number of surgically altered anatomy (SAA) patients with benign biliary obstruction per total number of patients in the series; ²Number of patients with attempted drainage by stent insertion and/or balloon dilation with/without stone removal (ITT); ³Number of patients with attempted stent placement (ITT), followed by number of metal stents in parentheses;

⁴Transmural variants of HG (hepatico-esophagostomy/jejunostomy) following intrahepatic access, except when stated otherwise;

⁵Number of successful cases or complications per number of SAA patients with benign biliary obstruction;

⁶Choledochoduodenostomy with plastic stents used for transmural stone extraction in Billroth I;

⁷Transpapillary AG metal stent placed at follow-up of prior transmural intrahepatic stenting in anastomotic Strx. Tallied in the total count;

⁸AGB prior to AGS in a single case. AGS tallied as main intervention and AGB not tallied in the total count;

⁹AGB with stone removal plus temporary NBD placement in four patients. AGB of anastomotic Strx and normal cholangiogram in one each. SAA: Surgically altered anatomy; ESCP: Endosonographic cholangiopancreatography; RYG: Roux-en-Y gastrectomy; RYGB: Roux-en-Y gastric bypass; RYHJ: Roux-en-Y hepatico-jejunostomy; RV: Rendezvous; AGS: Antegrade stent insertion; AGB: Antegrade balloon dilation; HG: Hepatico-gastrostomy; ITT: Intention-to-treat; Sn: Successful cases; Cn: Complications.

ESCP interventions for malignant biliary obstruction in SAA

Recurrent gastric, pancreatic and biliary malignancy after surgical resection usually present as obstructive jaundice. Roux-en-Y gastrectomy or hepatico-jejunostomy, pancreatico-duodenectomy and palliative bypass gastro-jejunostomy are the most commonly SAA encountered. PTBD has been classically used in this setting, as it is much simpler and more readily available than the combinations of surgical, percutaneous or enteroscopy-based ERCP described earlier. Two important limitations of PTBD in these patients are failure to provide internal biliary drainage occasionally[16] and the management of relapsing stent dysfunction[2].

As in other settings, EUSBD may overcome those limitations, improving management in patients with SAA and malignant biliary obstruction over that afforded by PTBD. Internal biliary drainage has been provided by hepaticogastrostomy (n = 10) or any of its variants (n = 7), with either plastic or metal stents placed in about half of the cases each (Table 6). Transmural metal stent placement carries the risk of severe bile leakage into the peritoneum caused by stent foreshortening[16]. This risk can be minimized by initial placement of transmural plastic stents followed by elective stent exchange after track maturation[2,14,16]. The recommended technique for plastic stent replacement after transmural EUSBD is over-the-wire snare removal after guide-wire cannulation of the stent[14,111]. The snare over-the-wire technique prevents track disruption and avoids the need for a repeat transhepatic (PTBD or EUSBD) puncture should stent occlusion lead to dysfunction.

As an alternative to transmural stenting, transductal (transpapillary or transanastomotic) stent placement has been reported in just five patients with SAA and malignant biliary obstruction, by means of rendezvous ERCP after antegrade guide-wire passage in 2 cases, and by direct antegrade stent insertion in a further 3 (Table 6). The final position of the stent in antegrade or rendezvous approaches is the same as in internal drainage by PTBD. Püspök et al[13] associated 7F plastic stent hepatico-jejunostomy in a patient with recurrent gastric cancer and prior Roux-en-Y total gastrectomy to antegrade transpapillary metal stent placement. Although their goal in placing this second transmural stent was to minimize the risk of acute leakage, this dual approach might well serve the purpose of avoiding a repeat transhepatic puncture in case of stent dysfunction, the transmural stent serving as an entry point to the bile-duct. A similar strategy to manage stent dysfunction was reported by Bories et al[16]. After initial transmural stenting, they converted drainage electively to antegrade stent placement at follow-up in two cases and to rendezvous in a further case. These authors highlight the ease of endoscopic re-intervention through a transmural metal stent in case of stent dysfunction.

ESCP interventions for benign biliary obstruction in SAA

Stones and anastomotic strictures are the most common benign causes requiring intervention after SAA, about half each in the 25 patients reported so far in the literature (Table 7). In contrast to SAA patients with malignant biliary obstruction in whom EUSBD is predominantly transmural, transductal intervention is the dominant approach to benign disease, chosen in 84% of SAA cases (Table 7). Antegrade balloon dilation with or without stone removal is the major ESCP intervention performed in this setting[57,110], with rendezvous[3,16,31,110] and antegrade stenting[109,112] having a secondary role.

Only two small series have specifically reported these EUSBD interventions in the setting of SAA. Weilert et al[110] succeeded at antegrade CBD stone removal in four RYGB patients and managed to salvage a further two with rendezvous enteroscopy-ERCP. These authors found rendezvous enteroscopy-ERCP with retrograde balloon dilation and stone removal helpful as a cross-over strategy following failed antegrade puncture track dilation. Iwashita et al[57] replicated their approach in patients with CBD stones following Roux-en-Y gastrectomy. To prevent leakage and to allow serial cholangiography, these authors left a nasobiliary drainage tube in place during 2-4 wk. Serial cholangiograms led to repeat intervention in one out of four patients with CBD stones.

Antegrade ESCP balloon dilation of anastomotic strictures without additional stenting has been reported as an effective measure to relieve biliary obstruction by several authors[34,109,113,114]. As effective remodeling of benign biliary strictures usually requires serial endotherapy[85,99], the long-term outcomes of single session balloon dilation remain unproven. A seemingly more effective two-step stricture remodeling strategy was used by Artifon et al[112], who via ESCP placed a temporary covered metal stent in an antegrade fashion across an anastomotic stricture. The stent was removed at follow-up using balloon enteroscopy. Authors limiting intervention to antegrade balloon dilation base the potential compromise in efficacy on concerns about transmural stenting when only minimal intrahepatic dilation is present[109], or about stent removability secondary to impaired access caused by SAA[114].