Although endoscopic biliary drainage for malignant hilar obstruction has been in use for the past 20 years, bilateral stent placement is still considered superior with respect to efficacy and complication rates. Supporters of unilateral drainage have emphasized that bilateral stent deployment is associated with a higher risk of complications 1 and that adequate palliation can be achieved by the drainage of only 25% of the liver 1–4. Nevertheless, many endoscopists advocate the use of endoscopic bilateral metal stents because unilateral drainage can be insufficient to relieve obstructive jaundice and has the potential to cause acute cholangitis in undrained ducts 5–8.
Most early complications of endoscopic bilateral stents, such as cholangitis, are associated with technical problems, including repeated guiding and excess occupation with the contrast media of the intrahepatic bile duct (IHD). Many new techniques have been used to decrease early complications 9–11, but the results are still insufficient. Furthermore, longer survival in patients with malignant hilar obstruction has allowed late complications, such as stent occlusion, to emerge. Percutaneous drainage has been carried out because of technical difficulties associated with the endoscopic approach. However, endoscopic revision provides more comfortable and physiologic conditions for patients; endoscopists have therefore attempted the use of various methods to overcome these technical difficulties 12–14.
To this end, we retrospectively investigated the technical and clinical efficacy of the ‘stent-in-stent’ method of implanting a ‘Y’ stent for malignant hilar obstruction. We evaluated the short-term and long-term complications associated with this method. Furthermore, we compared the outcome in patients who received bilateral stenting during the early period of the study with those treated later to adjust for any differences caused by newly designed stents and our increasing technical experience over the 5 years of this study.
Materials and methods
From August 2005 to July 2010, bilateral metal Y-stents were implanted in 97 patients with unresectable malignant hilar obstruction. Thirty-one patients received stents during the early period (August 2005 to May 2007) and 66 patients during the late period (June 2007 to July 2010) of the study. All patients had obstructive jaundice without acute septic cholangitis. The diagnosis was made principally on the basis of ultrasound, computed tomography, magnetic resonance cholangiopancreatography, and endoscopic retrograde cholangiography. The histological diagnosis of malignancy was established using transpapillary brushing cytology, forceps biopsy, or percutaneous fine-needle aspiration cytology. All patients were followed until death, except seven patients, who received unilateral metal stenting and were transferred to other hospice centers for terminal care. The institutional review board of our university approved this retrospective study.
The stents used in this study were self-expandable nitinol stents. The type of stents is the main variable to facilitate a comparison of clinical outcomes between the early and the late periods, which was concurrently changed at the starting time of the late period.
The transverse stent with the Y-configuration, called a Y-stent (8.5 Fr, Niti-S Biliary Y-stent; Taewoong, Seoul, Korea), was developed as a hybrid of spiral and Z-stents. The mesh of the central part of the Y-stent was made larger by omitting the Z component, resulting in a 10 mm-long open-weave portion located precisely at the center of the stent. During the late period of the study, the central part of the Y-stent (7.5 Fr, Niti-S Biliary Y-stent; MITech, Seoul, Korea) was extended to 2.5 cm in length, and had a relatively close-woven central portion compared with that of the Y-stent used during the early period (Fig. 1).
The second stents were a closed-cell design spiral or Z-stent (7.5 Fr, Niti-S or Niti-D stent, respectively; Taewoong) used during the early period, and a slimmer open-cell-design stent (7 Fr, Zilver Biliary Self-Expanding Stent; Cook Endoscopy, Winston-Salem, North Carolina, USA) used during the late period (Fig. 1).
Prophylactic treatment with broad-spectrum antibiotics was initiated before the procedure and was continued for 5–7 days thereafter. Contrast injections at the endoscopic retrograde cholangiopancreatography (ERCP) were deliberately limited to the extrahepatic bile duct distal to the tumor. Minor endoscopic sphincterotomy was carried out in all cases because bilateral stenting in patients with hilar cholangiocarcinoma requires multiple cannulations with many instruments, such as catheters, guidewires, dilator devices, and stents.
New techniques that were learned through experience were gradually applied throughout the entire study period. Selective cannulation of the IHDs was initially attempted using a conventional ERCP cannula, or a retrieval balloon catheter, and additionally by a pull-type sphincterotome that was used from the middle of the early period, or a triple-lumen catheter (Haber Ramp catheter; Cook Endoscopy) from the middle of the late period (Fig. 1). At the same time, guidewires were used through the catheters successively as follows: 0.025 inch-guidewire, 0.035 inch-guidewire, and a more hydrophilic Terumo guidewire (Terumo, Tokyo, Japan). After selective cannulation of both IHDs, the Y-stent was placed into the more inflectional side of the bile duct from the latter half of the early period, allowing better access to the contralateral side. Bilateral metal stent deployment was carried out as described previously by Kim et al. 15.
Stent occlusion was managed by mechanical cleaning alone with a biliary extraction balloon (Boston Scientific, Natick, Massachusetts, USA) and bilateral insertion of 7 Fr double-pigtail polyethylene plastic stents (Cook, Bloomington, Indiana, USA) through the lumen of the first Y-stent and the second metal stent, respectively (Fig. 1). If revision was endoscopically impossible or unsuitable, the obstructed bile duct was drained using percutaneous transhepatic biliary drainage (PTBD).
All endoscopic procedures were carried out by only one practiced endoscopist who performed more than 10 000 cases of ERCP and had placed more than 300 metallic stents for malignant biliary obstruction.
Definitions of events
Successful stent insertion was defined as passage of the stent across the stricture, along with the flow of contrast medium and/or bile through the stent. Successful drainage was defined as a decrease in bilirubin to less than 75% of the pretreatment value within the first month. Early and late complications were defined as those occurring within 30 days and after 30 days of stent placement, respectively. Stent occlusion was defined as the recurrence of jaundice and cholestasis and/or evidence of dilated biliary systems on a ultrasound or a computed tomography requiring biliary reintervention. Stent occlusion was classified as predominantly tumor ingrowth (seen as a narrowing within the stent on fluoroscopy) or tumor overgrowth (seen radiographically as new narrowing at the proximal or the distal margin of the stent). Stent patency was defined as the period between insertion and stent occlusion. The overall survival duration was defined as the period between the diagnosis of primary malignancy and death.
Main outcome measurements
The main outcome measurements are the technical success rate in terms of efficacy, cholangitis as an early complication, and cholecystitis as a late complication. In addition, we determined the functional success rate, procedure time, stent patency, and overall survival as parameters of efficacy. Cholecystits and pancreatitis as early complications and cholangitis and stent occlusion as late complications were evaluated.
Student’s t-test was used for continuous data comparison. Fisher’s exact probability test and Pearson’s χ 2-test were used for the comparison of categorical data as appropriate. Stent patency and patient survival time were assessed using the Kaplan–Meier method. A P value of less than 0.05 was considered significant. For analysis, SPSS software (version 11; SPSS Inc., Chicago, Illinois, USA) was used.
Of the 97 patients, 71 had primary cholangiocarcinoma, 23 had gallbladder carcinoma, two had periportal lymph node metastasis, and one had hepatocellular carcinoma accompanied by bile duct invasion. Using the Bismuth classification system, 24 cases were type II, 36 were type IIIA, 14 were type IIIB, and 22 were type IV. There were no significant differences in histologic type or Bismuth classification between patients in the early and the late treatment periods (Table 1).
Bilateral stent placement carried out using the ‘stent-in-stent’ method with a Y-stent was achieved in 76 of the 97 patients (technical success, 78.3%). The first Y-stent was placed successfully in all patients. Bilateral stents were implanted in 18 of 31 patients (58.1%) during the early period and in 58 of 66 patients (87.9%) during the late period. The rate of successful placement of bilateral stents was significantly higher during the late period than the early period (P=0.001). The success rate of selective cannulation of the contralateral IHD was 86.6% (84/97), and the rate during the late period was significantly higher than that during the early period (71.0%, 22/31 vs. 93.9%, 62/66; P=0.002). The second stent delivery system could pass the distal edge of the first Y-stent in 76 of the 84 patients (90.5%), and no statistical differences existed between the two groups (Table 2). Seven of the 21 patients who underwent unilateral metal stenting received contralateral drainage by stent insertion by a percutaneous transhepatic approach.
Of the 76 patients in whom technical success was achieved, 74 experienced successful drainage (functional success, 97.3%) as indicated by a significant decrease in the serum bilirubin level [median bilirubin levels: pretreatment, 7.92 mg/dl (range, 2.34–45.82 mg/dl); post-treatment, 1.07 mg/dl (range, 0.26–10.31 mg/dl); P<0.001].
Cholangitis as an early complication occurred in six patients (6.2%). During the early period, cholangitis occurred more frequently than during the late period (16.1%, 5/31 vs. 1.5%, 1/66; P=0.012). Cholecystitis and pancreatitis occurred in nine and two patients (9.3 and 2.1%), respectively, as early complications. Two patients died because of septic cholangitis and aspiration pneumonia within 7 days after stent insertion during the early period (Table 3).
Late complications occurred in 46 of 76 patients (60.5%). Cholecystitis occurred in 21 patients (27.6%) and was managed by PTBD. Cholangitis occurred in 12 patients (15.8%). Twenty-nine patients (38.2%) developed stent occlusion because of tumor recurrence. Among these 29 patients, 17 (58.6%) developed tumor ingrowth around the hilar portion and 12 (41.4%) developed tumor ingrowth and overgrowth in the hilar portion and more proximal (peripheral) intrahepatic duct. Of the 29 patients who developed stent occlusion, cholangitis occurred synchronously or metachronously in nine, cholecystitis in seven, and liver abscess in one as late complications. There was no significant difference in late complications between the early and the late periods (Table 3). The median period of stent patency was 159 days for all patients, with no statistical difference between the two groups (173 vs. 152 days; P=0.513; Fig. 2). However, the median period of stent patency in patients with cholangiocarcinoma was higher than that in patients with gallbladder carcinoma (191 vs. 87 days; P=0.004; Fig. 3).
Nine of the 29 (31.1%) patients who experienced stent occlusion were treated endoscopically by the insertion of bilateral plastic stents (7 Fr) into both ducts through bilateral metal stents, for which six patients received an open-cell-design stent as the second stent. Eleven patients (37.9%) were treated through unilateral plastic stent insertion: five of these patients received additional PTBD in the contralateral bile duct because of cholangitis. Six patients (20.7%) with multiple strictures because of tumor ingrowth and overgrowth were not suitable for endoscopic treatment and were instead treated with PTBD. Three patients (10.3%) with septic cholangitis were treated with PTBD (Table 4).
The median follow-up period was 234 days (range, 10–824 days). The median survival was 226 days in all patients, with no statistical difference between the early and the late periods (240 vs. 212 days; P=0.282; Fig. 2). Furthermore, there were no statistically significant differences between cholangiocarcinoma and gallbladder carcinoma (253 vs. 166 days; P=0.065; Fig. 3).
Malignant biliary obstruction results in pruritus, pain, jaundice, and cholangitis, often decreasing the quality of life. For unresectable cases, endoscopic stenting of biliary obstruction is considered to be the preferred palliative modality because it is a less invasive, more comfortable, and more physiological procedure than either the percutaneous approach or surgical bypass 12,16–19. Unilateral stenting can often provide adequate management of jaundice or cholangitis 1–4, but the necessity of bilateral stent deployment has been discussed in relation to patients with Bismuth III or IV stricture because of an increased risk of cholangitis in an undrained duct that had become filled with the contrast media 5–8.
Since Silverman and Slivka 18 initially described the technique of partial stent-in-stent deployment, this technique has been used as an alternative endoscopic procedure that allows bilateral metal stenting to be placed in a Y-configuration in which a second stent traverses the open-mesh wall of the first stent to enter the contralateral biliary duct. However, several technical and clinical problems were encountered in the use of bilateral metal stenting with a partial stent-in-stent technique. First, it was occasionally very difficult to place the second stent on the already occupied contralateral biliary system. Second, repeated cannulation of the IHD could cause early procedure-related cholangitis. Third, when stent occlusion occurred, endoscopic revision was difficult even if a hydrostatic balloon or a 7 Fr Soehendra stent extractor was used. Fourth, the occurrence of cholecystitis increased as a late complication because of obstruction of the cystic duct orifice because of the tightly woven mesh composition of the two overlapping stents and the aggravation of the tumor mass around the hilar portion.
To overcome these limitations, we used the following devices and techniques. First, the first stent was placed in the more inflectional side of the bile duct for successful access to the contralateral side. Second, a newly designed Y-stent with a longer central mesh segment allowed better placement of the central portion of the stent in the hilar bifurcation. The relatively close-woven central portion of this stent enhanced the radial force, resulting in maintenance of the central mesh spaces and creation of a larger diameter of the contralateral duct. As a result, selective cannulation of the contralateral biliary duct was facilitated. Third, the guidewires through the middle and distal opening of the triple-lumen catheter or the distal end of the rotatable ERCP cannula could be deflected off the lateral wall of the common hepatic duct or the common bile duct toward the desired ducts. Fourth, we used a slimmer, open-cell stent as the second stent (8.5–>7.0 Fr) to help the second stent pass more easily through the tight central mesh of the first stent; endoscopic revision was easier to carry out than closed-cell stents, which have a strong restoring force after dilation.
Our 78.3% technical success rate was lower than those reported in recent studies (85.3–100%; Table 5) 15,20–25, which were comparable with our 87.9% (58/66) technical success rate during the late period. Many researchers have described new devices and various techniques for bilateral metallic stents placement across hilar malignancy, including stent-by-stent parallel placement 26,27, a partial stent-in-stent method using large-cell-width stents or open-cell stents 28, and a temporary plastic stent to facilitate the placement of bilateral metal stents 29. As a result, high technical success rates have been reported in more recent studies 15,20–29.
Early complications including cholangitis and cholecystitis occurred in 16 of 97 patients (16.5%) in our study. Recent reports have documented an incidence in early complications of 0–11.1% because of a decreased rate of cholangitis (0–3.3%) 15,20–25. The incidence of cholangitis decreased in the late period implants compared with that in the early period (16.1–1.5%). We believe that these results were because of the process of implantation, which required relatively fewer procedures. However, the incidence of cholecystitis was not altered (9.7–9.1%) because of the technical success achieved in the repair of severe hilar obstruction with the occlusion of the cystic duct and impaired gallbladder emptying.
Cholecystitis occurred as a late complication in 21 of 76 patients (27.6%), which was more frequent than that in other recent studies (6.1–6.9%) 15,20–25. Possible explanations for this result are as follows: (a) the cystic duct was obstructed because of the tightly woven mesh composition of the two overlapping stents and the aggravation of tumor mass around the hilar portion, (b) the relatively large number of patients (25) with gallbladder cancer who were enrolled in our study, and (c) because the enrolled patients were followed over a comparatively longer period of time.
Management of metal stent occlusion can be technically more difficult when stents are placed in hilar lesions, especially when bilateral stents are placed using the partial stent-in-stent technique. In our study, incidence rates of stent occlusion were similar between the early and the late periods (42.1 vs. 34.2%, respectively). Although the difference was not statistically significant, eight of 22 patients (36.4%) were treated endoscopically with bilateral plastic stenting during the late period compared with 14.3% during the early period (one of seven patients). Endoscopic revision including unilateral plastic stenting was achieved in 15 patients (51.7%), which is comparable with recently reported results (40.0–73.3%) 15,20–25.
Finally, the overall median survival was 226 days in our study. Considering the relatively large number of patients with gallbladder cancer, our result was comparable with that observed in a previous study (180–260 days) 15,20–25. Although there was no statistical significance, the overall median survival of patients with cholangiocarcinoma was longer than that of patients with gallbladder carcinoma (253 vs. 166 days; P=0.065). We believe that this result indicates the significantly longer stent patency of patients with cholangiocarcinoma compared with that of patients with gallbladder carcinoma (191 vs. 87 days; P=0.004).
Our study is limited by its retrospective, single-center design. We had difficulty distinguishing the study periods because the introduction of new devices and new techniques was not concurrent. However, the long central mesh Y-stent and the slimmer, open-cell second stent were concurrently changed at the starting time of the late period, which led to the main differences between the two groups, although each group represented technical diversity. As a result, this influence was considered relatively small. Furthermore, the relatively large number and the longer follow-up period of enrolled patients could counterbalance a selection bias in this retrospective study. Prospective randomized studies are warranted in the future.
Endoscopic bilateral metal stenting using a Y-stent appears to be an attractive palliative method for patients with malignant hilar obstruction. Many technical and clinical problems related to the partial stent-in-stent technique are likely to be overcome with the development of new devices and the accumulation of technical experience.
This study was supported by a grant of the Korea Healthcare technology R&D Project, Ministry for Health, Welfare & Family Affairs, Republic of Korea (A091047).
Conflicts of interest
There are no conflicts of interest.
1. Palma GD, Galloro G, Siciliano S, Iovino P, Catanzano C. Unilateral versus bilateral endoscopic hepatic duct drainage in patients with malignant hilar biliary obstruction: results of a prospective, randomized, and controlled study. Gastrointest Endosc. 2001;53:547–553
2. Peters RA, Wiliams SG, Lombard M, Karani J, Westaby D. The management of high-grade strictures by endoscopic insertion of self-expanding metal endoprostheses. Endoscopy. 1996;28:10–16
3. Freeman ML, Overby CRE. Selective MRCP and CT-targeted drainage of malignant hilar biliary obstruction with self-expanding metallic stents. Gastrointest Endosc. 2003;58:41–49
4. Palma GD, Pezzullo A, Rega M, Persico M, Pantrone F, Mastantuono L, Persico G. Unilateral placement of metallic stents for malignant hilar obstruction: a prospective study. Gastrointest Endosc. 2003;58:50–53
5. Polydorou AA, Chisholm EM, Romanos AA, Dowsett JF, Cotton PB, Hatfield AR, Russell RCG. A comparison of right versus left hepatic duct endoprosthesis insertion in malignant hilar biliary obstruction. Endoscopy. 1989;21:266–271
6. Chang WH, Kortan P, Haber GB. Outcome in patients with bifurcation tumors who undergo unilateral versus bilateral hepatic duct drainage. Gastrointest Endosc. 1998;47:354–362
7. Sherman S. Endoscopic drainage of malignant hilar obstruction: is one biliary stent enough or should we work to place two? Gastrointest Endosc. 2001;53:681–684
8. Kogure H, Isayama H, Kawakubo K, Sasaki T, Yamamoto N, Hirano K, et al. Endoscopic bilateral metallic stenting for malignant hilar obstruction using newly designed stents. J Hepatobiliary Pancreat Sci. 2011;18:653–657
9. Ching CK, Lai KC, Hu W, Lam SK. Cannulatome-aided selective intrahepatic bile duct cannulation. Gastrointest Endosc. 1996;43:632–633
10. Moxon DR, Hong K, Brown RD, Venu RP. Selective intrahepatic ductal cannulation during ERCP with a sphincterotome. Gastrointest Endosc. 2003;57:738–743
11. Kim JY, Kang DH, Choi CW, Kim HW, Park SB, Kim DU. Selective intrahepatic duct cannulation by using a triple-lumen catheter for endoscopic bilateral stenting in hilar cholangiocarcinoma. Gastrointest Endosc. 2010;72:192–198
12. Ferrucci JT, Mueller PR, Harbin WP. Percutaneous transhepatic biliary drainage. Radiology. 1980;135:1–13
13. Tham TC, Carr-Locke DL, Vandervoort J, Wong RC, Lichtenstein DR, Van Dam J, et al. Management of occluded biliary Wallstents. Gut. 1998;42:703–707
14. Ridtitid W, Rerknimitr R, Janchai A, Kongkam P, Treeprasertsuk S, Kullavanijaya P. Outcome of second interventions for occluded metallic stents in patients with malignant biliary obstruction. Surg Endosc. 2010;24:2216–2220
15. Kim JY, Kang DH, Kim HW, Choi CW, Kim ID, Hwang JH, et al. Usefulness of slimmer and open-cell-design stents for endoscopic bilateral stenting and endoscopic revision in patients with hilar cholangiocarcinoma. Gastrointest Endosc. 2009;69:112–125
16. Soehendra N, Frederix VR. Palliative bile duct drainage – a new endoscopic method of introducing a transpapillary drain. Endoscopy. 1980;12:8–11
17. Polydorou AA, Cairns SR, Dowsett JF, Hatfield ARW, Salmon PR, Cotton PB, et al. Palliation of proximal malignant biliary obstruction by endoscopic endoprosthesis insertion. Gut. 1991;32:685–689
18. Silverman W, Slivka A. New technique for bilateral metal mesh stent insertion to treat hilar cholangiocarcinoma. Gastrointest Endosc. 1996;43:61–63
19. Dumas R, Demuth N, Buckley M, Peten EP, Manos T, Demarquay F, et al. Endoscopic bilateral metal stent placement for malignant hilar stenoses: identification of optimal technique. Gastrointest Endosc. 2000;51:334–338
20. Kawamoto H, Tsutsumi K, Harada R, Fujii M, Kato H, Hirao K, et al. Endoscopic deployment of multiple JOSTENT SelfX is effective and safe in treatment of malignant hilar biliary strictures. Clin Gastroenterol Hepatol. 2008;6:401–408
21. Park DH, Lee SS, Moon JH, Choi HJ, Cha SW, Kim JH, et al. Newly designed stent for endoscopic bilateral stent-in-stent placement of metallic stents in patients with malignant hilar biliary strictures: multicenter prospective feasibility study. Gastrointest Endosc. 2009;69:1357–1360
22. Hwang JC, Kim JH, Lim SG, Kim SS, Yoo BM, Cho SW. Y-shaped endoscopic bilateral metal stent placement for malignant hilar biliary obstruction: prospective long-term study. Scand J Gastroenterol. 2011;46:326–332
23. Chahal P, Baron TH. Expandable metal stents for endoscopic bilateral stent-within-stet placement for malignant hilar biliary obstruction. Gastrointest Endosc. 2010;71:195–199
24. Iwano H, Ryozawa S, Ishigaki N, Taba K, Senyo M, Yoshida K, et al. Unilateral versus bilateral drainage using self-expandable metallic stent for unresectable hilar biliary obstruction. Dig Endosc. 2011;23:43–48
25. Kanno Y, Ito K, Fujita N, Noda Y, Kobayashi G, Obana T, et al. Single-session endoscopic bilateral y-configured placement of metal stents for hilar malignant biliary obstruction. Dig Endosc. 2011;23:91–96
26. Saleem A, Baron TH, Gostout C. Large-diameter therapeutic channel duodenoscope to facilitate simultaneous deployment of side-by-side self expandable metal stents in hilar cholangiocarcinoma. Gastrointest Endosc. 2010;72:628–631
27. Chennat J, Waxman I. Initial performance profile of a new 6F self-expanding metal stent for palliation of malignant hilar biliary obstruction. Gastrointest Endosc. 2010;72:632–636
28. Kogure H, Isayama H, Nakai Y, Tsujino T, Ito Y, Yamamoto K, et al. Newly designed large cell Niti-S stent for malignant hilar biliary obstruction: a pilot study. Surg Endosc. 2011;25:463–467
29. Hookey LC, Moine OL, Deviere J. Use of a temporary plastic stent to facilitate the placement of multiple self-expanding metal stents in malignant biliary hilar strictures. Gastrointest Endosc. 2005;62:605–609
© 2013 Lippincott Williams & Wilkins, Inc.