Endoscopic biliary drainage for distal bile duct obstruction due to pancreatic cancer

1) Stents for transpapillary drainage

Transpapillary drainage can be of two types as follows: endoscopic biliary stenting, which involves an internal fistula; and endoscopic nasobiliary drainage, which involves an external fistula. Stents include plastic stents (PSs) and self-expanding metal stents (SEMSs), and SEMS include uncovered SEMS (USEMSs) and covered SEMS (CSEMSs). The axial force (AF) and radial force (RF) are significant factors affecting the mechanical properties of SEMSs.6 The AF is the straightening force on the central axis of the stent, and the RF is the expanding force that induces the radial deformation of the stent. The AF enhances the flexibility of the stent within the bile duct, whereas the RF promotes bile duct dilation. In addition, the function of the biliary stent and the occurrence of AEs are closely related to both the AF and RF. A strong AF can lead to acute cholecystitis and pancreatitis after postendoscopic retrograde cholangiopancreatography.7 A weak RF is associated with stent dislocation.8 Recently, antireflux metal stents (ARMSs) and large- or small-diameter SEMSs have become commercially available, and their usefulness has been reported.

2) PS vs. SEMS

Transpapillary biliary drainage stents are broadly divided into PSs and SEMSs. Zorrón Pu et al.9 reported a meta-analysis comparing PS and SEMS placement for unresectable distal MBO. In this meta-analysis, SEMS placement was associated with lower overall stent dysfunction (21.6% vs. 46.8%, p<0.00001) and fewer reinterventions (21.6% vs. 56.6%, p<0.00001), with no difference in complications (13.7% vs. 15.9%, p=0.16) compared to PS placement. Moreover, the mean overall survival was higher in the SEMS group (182 vs. 150 days, p<0.0001) than in the PS group. In general, SEMSs are preferable to PSs for treating unresectable malignant distal biliary obstructions. In two randomized controlled trials (RCTs) limited to unresectable malignant distal biliary obstruction in pancreatic cancer,10,11 SEMSs showed lower overall stent dysfunction than PSs, with no difference in complication rates. The results of these RCTs suggest that SEMSs, rather than PSs, should be used for unresectable malignant distal biliary obstructions in pancreatic cancer.

3) USEMS vs. CSEMS

Although a USEMS provides a large-diameter lumen, its mesh structure has the disadvantage of tumor ingrowth, where tumors invade the lumen through the gaps between the meshes, causing the stent to become occluded. A CSEMS, a membrane-covered SEMS designed to prolong patency, was developed to overcome this disadvantage. Several RCTs comparing USEMSs and CSEMSs have been reported (Table 1).10,12-18

Table 1.

Randomized controlled trials comparing USEMS and CSEMS for unresectable distal biliary obstruction

In 2004, Isayama et al.12 reported that in an RCT comparing CSEMSs and USEMSs for distal biliary obstruction in malignant tumors, the recurrent biliary obstruction (RBO) rate of CSEMSs was 86% and the median time to recurrent biliary obstruction (TRBO) was 304 days, whereas USEMSs showed an RBO rate of 62% and a median TRBO of 161 days. CSEMSs are superior to USEMSs in terms of cumulative patency rate and patency duration. However, CSEMSs have drawbacks such as stent deviation (migration) and cholecystitis. An RCT reported by Kullman et al.13 in 2010 showed a median TRBO of 152 days for CSEMSs and 199 days for USEMSs, with no significant differences between the two stent types.

In 2013, Kitano et al.15 reported the results of a multicenter RCT comparing partially covered SEMSs (PCSEMSs), which have an uncovered flare on the biliary side of the covered stent, to prevent migration, with USEMSs. The study showed that PCSEMSs resulted in a significantly longer TRBO (219 vs. 166 days) than USEMSs. Unlike studies that analyzed patients with distal biliary obstruction caused by various conditions such as pancreatic, biliary, and gastric cancers, this study was limited to distal biliary obstruction caused by pancreatic cancer, thereby demonstrating the superiority of CSEMSs in pancreatic cancer. Furthermore, migration was not observed in the PCSEMS group. In a 2018 multicenter RCT by Conio et al.,17 in which 74% of the patients had pancreatic cancer, the median TRBO was significantly shorter in the CSEMS group than in the UCSEMS group (240 vs. 541 days). A meta-analysis by Yamashita et al.19 in 2022 reported no difference in the RBO rate between the USEMS and CSEMS groups in the entire cohort; however, when the analysis was limited to patients with pancreatic cancer, the CSEMS group had a significantly lower RBO rate. In addition, PCSEMSs specifically designed to prevent migration were associated with a significantly longer TRBO than UCSMSs, which was not observed in other CSEMSs. No significant differences were observed in procedure-related AEs, including migration, between the USEMS and CSEMS groups. Whether USEMS or CSEMS is recommended for patients with pancreatic cancer and unresectable malignant distal biliary obstruction remains to be determined. However, as shown in the meta-analysis, CSEMSs with antimigration properties had a longer TRBO than USEMSs. Moreover, CSEMSs can prevent stent occlusion as long as the membrane is not damaged, and reintervention is relatively easy with stent replacement, as in the case of PSs. The Japanese guidelines for the treatment of pancreatic cancer recommend the use of CSEMSs for unresectable malignant distal biliary obstruction.20

4) ARMS

ARMSs were developed to prevent RBO and reflux cholangitis caused by the reflux of intestinal contents and food impaction, which are among the causes of RBO after stent placement. Although various shapes of antireflux valves have been reported, mainly from Asia, two types of ARMSs are available: the chimney and duckbill types.21,22 Currently, there is no clear evidence that ARMSs prolong the TRBO. However, in patients in whom the cause of the initial stent occlusion is reflux cholangitis or in patients prone to reflux cholangitis, such as those with duodenal stenosis, ARMSs can be useful22 (Fig. 1) and should be considered an option.

Fig. 1.

A case of repeated reflux cholangitis after fully covered self-expandable metal stent placement and duodenal stenting for distal biliary obstruction and duodenal obstruction. (A) Duckbill biliary stent (Kawasumi Laboratories Inc.): The stent has two additional meshes at the distal end to regulate the opening of the valve; the valve is normally closed to prevent reflux of duodenal contents into the bile duct, but opens to allow bile to flow out when the bile duct pressure increases. (B) Radiographic image: Duodenography showing reflux of the contrast medium into the bile duct. (C) Radiographic image: After placement of the Duckbill biliary stent, duodenography shows no reflux of the contrast medium into the bile duct.

5) Large-diameter SEMS

Although 10-mm diameter SEMSs are commonly used, 12- and 14-mm diameter SEMSs have been developed to prolong the TRBO. In a prospective comparative study by Lee et al.23 comparing 10-mm and 12-mm diameter SEMSs, the RBO rate was significantly lower in the 12-mm SEMS group (RBO: 8.0% in the 12-mm diameter SEMS group and 23.1% in the 10-mm diameter group), and the TRBO was significantly longer in the 12-mm SEMS group than in the 10-mm SEMS group (228 vs. 157 days, respectively). In addition, because the 12-mm SEMSs used in this study were equipped with an antimigration feature, the rate of stent migration was lower in the 12-mm group than in the 10-mm group (8% vs. 23.1%, p=0.036). In a cohort study by Satoh et al.,24 the TRBO was significantly longer in the 14-mm diameter SEMS group than in the 10-mm diameter group (median, 290 days for 10-mm diameter vs. not reached for 14-mm diameter). Further evaluation of large-diameter SEMSs vs. standard-diameter SEMSs in RCTs is needed.

6) Small-diameter SEMS

Six- and eight-millimeter diameter SEMSs were developed to reduce AEs. In a cohort study by Kawashima et al.25 comparing 8-mm and 10-mm SEMSs, the 8-mm diameter SEMSs were noninferior to the 10-mm diameter SEMSs. The survival time, incidence of AEs, and non-RBO rate at the time of death did not differ significantly between the two groups. In a cohort study by Harai et al.26 comparing 6-mm and 10-mm diameter SEMSs, the RBO rate was comparable between the two diameters, and the 6-mm SEMSs had fewer stent-related AEs. Further evaluation of small-diameter SEMSs vs. standard-diameter SEMSs in RCTs is needed.

1) Current status of preoperative drainage

In 2010, van der Gaag et al.27 reported the results of an RCT showing that preoperative drainage was not necessary in patients with MBO and serum total bilirubin levels <14.6 mg/dL. However, this trial was associated with problems such as a higher rate of drainage-related AEs than is usually observed, and preoperative drainage continues to be used in clinical practice. Moreover, with the development of chemotherapy and radiotherapy, neoadjuvant therapy (NAT), including NAC or neoadjuvant chemoradiotherapy, has attracted attention for its use in resectable and borderline resectable pancreatic cancer because of its ability to reduce tumor size and increase R0 resection rates.28-30 The time to surgery was longer with NAT than with upfront surgery. Therefore, the preoperative drainage period is longer, and the risk of cholangitis associated with stent occlusion due to chemotherapy is increased. SEMSs, which are associated with longer TRBOs, are used more frequently, whereas PSs used to be the preferred method for short-term preoperative drainage.

2) Stent selection in preoperative drainage31-35

Mandai et al.32 conducted an RCT comparing PSs with CSEMSs in patients with pancreatic cancer who underwent upfront surgery. Although the RBO rate was significantly lower in the CSEMS group, the intraoperative bleeding and surgery-related AEs were significantly higher in the CSEMS group than in the PS group. In contrast, an RCT of PSs versus CSEMSs in patients with borderline resectable pancreatic cancer scheduled for NAC conducted by Tamura et al.33 showed that the RBO rate was significantly lower in the CSEMS group, and there was no difference in intraoperative bleeding or other surgery-related AEs between the two groups. A meta-analysis by Watanabe et al.36 reported that SEMSs reduced the RBO rate and cost compared to PSs in the preoperative drainage of pancreatic cancer and reduced postoperative AEs in high-volume centers. Although preoperative drainage with SEMSs has a lower rate of RBO than that with PSs, the impact on surgery-related AEs requires further data collection. Two RCTs comparing the results of CSEMSs and USEMSs for preoperative drainage34,35 showed no significant difference in RBO rates or AEs between drainage using CSEMSs and USEMSs. Further data on the efficacy of CSEMSs and USEMSs for preoperative drainage are required (Table 2).

Table 2.

Randomized controlled trials analyzing stent selection in preoperative drainage for pancreatic cancer

1) Indications for EUS-BD

In a meta-analysis of EUS-BD that included 42 references, Wang et al.38 reported success, clinical success, and AE rates of 94.7%, 91.6%, and 23.3%, respectively. In this study, the most common AEs were bleeding, bile leakage, and stent migration. In particular, bile leakage and stent migration are potentially serious AEs. EUS-BD is currently indicated in cases of transpapillary biliary drainage failure or inability and has been compared with PTBD, which was originally performed. A meta-analysis reported that EUS-BD is significantly superior to PTBD in terms of clinical efficacy, as well as in AE and RBO rates, although the technical success rate is comparable between the two methods.39 EUS-BD creates an internal fistula, whereas PTBD creates an external fistula, and EUS-BD is superior in terms of the quality of life. Dhir et al.40 compared EUS-BD (EUS-RV) and precut in cases of difficult transpapillary biliary drainage and reported no difference in the rate of AEs between the two groups (EUS-BD, 3%; precut, 7%); however, EUS-BD was significantly superior in terms of the technical success rate (EUS-BD, 98%; precut, 90%). Therefore, EUS-BD may not only be the first choice as an alternative treatment in cases of failure or incapability of biliary drainage but may also be used as a treatment strategy at an early stage or for first drainage in cases where biliary drainage is expected to be difficult.

2) Results and comparison of EUS-BD techniques

Among the EUS-BD techniques, EUS-HGS and EUS-CDS with fistula formation have shown technical and clinical success rates of >90%, with no significant differences between EUS-HGS and EUS-CDS (Table 3).41-47 An RCT comparing EUS-HGS with EUS-CDS in malignant distal obstruction reported technical success rates of 87.5% and 82.6% and clinical success rates of 100% and 94.7% in the HGS and CDS groups, respectively. In addition, the AE rate, stent patency, and survival time did not differ between the groups.46 A meta-analysis conducted in 2022 reported comparable procedural and clinical success rates (EUS-HGS: 95% and 93.1%, respectively; EUS-CDS: 96.6% and 91.3%, respectively).48 On the other hand, EUS-CDS has a shorter procedure time (EUS-HGS, 35.4 minutes; EUS-CDS, 33.8 minutes) and a lower rate of complications (EUS-HGS, 19%; EUS-CDS, 13%) than EUS-HGS. Although further multicenter RCTs are required to compare the efficacy of EUS-HGS and EUS-CDS, readily switching from one method to another can increase technical success rates when one procedure is particularly challenging.46

Table 3.

Prospective studies and RCTs comparing EUS-CDS and EUS-HGS

EUS-AGS was first reported by Nguyen-Tang et al. in 2010,49 but EUS-AGS has problems such as bile leakage from the puncture site, and reintervention in the event of RBO is difficult. To solve these problems, EUS-HGAS (EUS-AGS+HGS), in which a stent is simultaneously placed in the puncture route in addition to EUS-AGS, has been developed and is now the mainstream EUS-AGS technique for MBO.50,51 Ishiwatari et al.52 designed a cohort study comparing EUS-HGS and EUS-HGAS for malignant distal biliary obstruction and reported that EUS-HGAS was superior regarding the TRBO (EUS-HGAS, not reached vs. EUS-HGS, 234 days). Although the procedure has some drawbacks, such as the need for a guidewire to pass through the biliary structure and the risk of acute pancreatitis when the stent is deployed across the papilla, EUS-HGAS is useful in cases of malignant distal biliary obstruction.

The first report on EUS-RVs was published by Mallery et al. in 2004.53 EUS-RV provides access to the biliary ductal system via multiple routes, creating only a temporary fistula. A literature review of 15 published articles, including 382 patients, reported that the overall success rate of the EUS-RV was 81%, and the AE rate was 10%.54 The success rates for the intrahepatic and extrahepatic bile duct puncture routes were 65% and 87%, respectively. Major complications included bleeding, bile leakage, peritonitis, pneumoperitoneum, and pancreatitis.

3) Stents used for EUS-BD

Various stent types have been used in EUS-BD. Initially, PSs were used, but SEMSs, such as biliary metal stents, were introduced later. Finally, lumen-apposing metal stents (LAMSs), specifically designed for EUS-guided drainage and providing anchorage across luminal structures, have become available.

(1) EUS-HGS

PS

For EUS-BD, double-pigtailed, single-pigtailed, or 7 to 10 Fr straight stents are used. PSs are smaller in diameter and carry a risk of RBO but are less expensive. Recently, single pigtail stents developed specifically for interventional EUS have become available, and excellent outcomes have been reported with these stents (Through & Pass, Type IT; Gadelius Medical K.).55 The tapered tip and straight distal side of this stent improve stent insertion during EUS-BD, and the four flanges and pigtail anchors prevent proximal and distal migration of the stent, respectively.

SEMS

Several tubular biliary stents have been developed for EUS-BD. However, these SEMS delivery systems are large in diameter and usually require tract dilatation during EUS-BD. SEMS specifically designed for EUS-BD vary in terms of design (braided or laser-cut), extent of coverage (uncoated, partially coated, or fully coated), antimigration properties, and delivery system diameter.

Among braided SEMSs, the Niti-S S-type stent (Taewoong Medical) is a PCSEMS with a 1 cm uncoated portion on the biliary side.56 This stent has a weak hold on the gastrointestinal wall and no antimigrational properties on the luminal side, which may cause delayed migration out of the gastrointestinal tract into the abdominal cavity. To overcome this drawback, modified stents have been developed, such as the Niti-S Spring Stopper Stent (Taewoong Medical) (Fig. 2) and the Hybrid Bonastent (Standard Sci. Tech) and the GIOBOR stent (Taewoong Medical), which has antimigrational properties on the luminal side.57-59 However, the diameter of these SEMS delivery systems is 8.0 to 8.5 Fr, and tract dilatation is usually required during EUS-BD.

Fig. 2.

Endoscopic ultrasound-guided hepaticogastrostomy using the Spring Stopper Stent (Taewoong Medical). (A) Partially braided self-expandable metal stent with a spring-type stopper on the luminal side (Niti-S Spring Stopper Stent; Taewoong Medical). (B) Radiographic image: fistula formation using a Spring Stopper Stent between the intrahepatic bile duct and stomach. (C) Endoscopic view of the Spring Stopper Stent.

Among the laser-cut SEMSs, the Covered BileRush Advance (Piolax Medical Devices) is a PCSEMS with a 1 to 2 cm uncovered area on the biliary side.60-62 This stent has a delivery system for a 0.025-inch guidewire with a 2.4 Fr tip and a 7 Fr shaft for direct insertion without tract dilatation. The stent is laser-cut, making it highly effective in holding down the gastrointestinal tract walls. In addition, this stent has a flared structure on both sides, which decreases the rate of delayed migration. Another laser-cut SEMS under development is the Hook Stent (ZEON Medical), which is a PCSEMS with a 1 cm uncoated portion on the luminal side (Fig. 3).63,64 This stent can be inserted directly without tract dilatation using a delivery system with a 7.2 Fr shaft and a hook structure on the luminal side to prevent migration.

Fig. 3.

Endoscopic ultrasound-guided hepaticogastrostomy using the Hook Stent (ZEON Medical). (A) Partially laser-cut self-expandable metal stent with a hook structure on the luminal side (Hook Stent). (B) A thin delivery system for a 0.025-inch guidewire with a 7.2-Fr shaft. (C) Endoscopic view of the Hook Stent.

PS vs. SEMS

In a multicenter cohort study of EUS-BD, including EUS-HGS and EUS-CDS, the TRBO of SEMSs and PSs were 72 and 97 days, respectively.65 Kanno et al.66 also reported that the TRBO of SEMSs did not differ from that of PSs (metal stents, 339 days; PSs, 125 days; p=0.61). In contrast, Hashimoto et al.67 reported that the median TRBO in EUS-HGS (SEMSs: 313 days; PSs: 125 days; p=0.01) in the MS group was longer than that in the PS group, whereas the median TRBO of SEMSs in EUS-CDS did not differ from that of PSs (SEMSs: not reached; PSs: 58 days; p=0.11). In EUS-HGS, the TRBO of SEMSs may be longer than that of PSs, although a larger RCT will be needed in the future.

4) Reintervention after EUS-BD

Because advances in chemoradiotherapy and immunotherapy for underlying malignancies have extended survival, the rate of RBO during the follow-up period has increased. This implies an increasing clinical requirement for reintervention following EUS-BD. In a multicenter cohort study by Minaga et al.75 on endoscopic reintervention after SEMS placement for EUS-BD, the technical and clinical success rates of endoscopic reintervention were 100% and 81.8%, respectively. Endoscopic reintervention after EUS-BD is technically feasible and clinically effective without severe AEs, especially in patients with distal MBOs.

1) EUS-BD for first drainage

The optimal indication for EUS-BD is obstructive jaundice due to unresectable MBO after failure of transpapillary biliary drainage. However, a recent RCT comparing transpapillary biliary drainage and EUS-BD as the first drainage procedure showed that the efficacy and safety of EUS-BD were comparable to those of transpapillary biliary drainage.76,77 Chen et al.78 conducted a multicenter RCT comparing transpapillary biliary drainage and EUS-CDS using LAMSs as the first drainage procedure. These results indicated that EUS-CDS using LAMSs is a safe and efficient first-line treatment option for patients with malignant distal biliary obstruction. The advantages of EUS-BD over transpapillary biliary drainage include the absence of stent occlusion due to tumor ingrowth and the absence of pancreatitis, as the drainage route does not include the tumor and papilla. However, serious AEs, such as bile leakage during the procedure and biliary peritonitis due to stent migration after EUS-BD, are important problems. Recent advances in the devices used in EUS-BD and technical improvements have reduced the incidence of these AEs. Advances in this device include the development of SEMSs that are specific for EUS-BD. Several types of metal stents designed specifically for EUS-BD have been developed recently and are reported to be effective in patients undergoing biliary drainage.79 Metal stents with thin delivery systems can be inserted directly, thereby skipping the tract dilatation step, which is expected to reduce bile leakage during EUS-BD.60,62 Technical improvements in EUS-BD include bile aspiration during EUS-BD, the double-guidewire technique, and the intrascopic channel stent release technique. Bile aspiration during EUS-BD decreases the pressure in the bile duct and bile leakage.52 The double-guidewire technique, in which two guidewires are inserted through the needle tract, results in an obtuse guidewire angle at the insertion site and provides scope stabilization for smooth device insertion.80 The intrascopic channel stent release technique, in which the EUS scope is stabilized until the stent is deployed up to 1 cm within the EUS scope, shortens the distance between the hepatic parenchyma and the gastric wall, thereby reducing bile leakage and stent migration.81 Owing to these dedicated devices and technical tips, EUS-BD may replace transpapillary biliary drainage as the first choice treatment for MBO.

2) EUS-BD in preoperative drainage

Currently, transpapillary biliary drainage is used as preoperative drainage, and PTBD is performed in cases of duodenal stenosis or unsuccessful cannulation. However, as mentioned above, once NAC has been performed, it is difficult to continue external fistula drainage from a quality-of-life perspective. Mukai et al.82 performed EUS-HGSs for preoperative drainage and reported no intraoperative or postoperative stent-related AEs. In addition, a multicenter cohort study by Janet et al.83 compared the technical and clinical success rates of EUS-CDS and transpapillary drainage for pancreatoduodenectomy after preoperative drainage. In the transpapillary drainage group, the technical and clinical success rates were 83.5% and 70.2%, respectively, whereas those in the EUS-CDS group were 100% and 89.3%, respectively (p=0.02 and p=0.05, respectively). Furthermore, the overall complication rate after pancreatoduodenectomy was higher in the transpapillary group than in the EUS-CDS group (92.2% vs. 75.0%, p=0.016). However, evidence supporting EUS-BD as a preoperative drainage technique is lacking, and more data should be collected in the future.