World J. Surg. 16, 1066-1073, 1992
World Journal of Surgery © 1992 by the Soci~t~ lnternationale de Chirurgie
Endoscopic Therapy for Biliary Obstruction Nib Soehendra, M.D., K e n n e t h F. Binmoeller, M.D., and H o r s t Grimm, M.D. Department of Endoscopic Surgery, University Hospital of Hamburg, Hamburg, Federal Republic of Germany Endoscopic management of biliary obstruction is feasible in most patients and has emerged as standard treatment. Aside from the removal of bile duct stones, placement of a biliary stent is the most commonly employed modality of management. In experienced hands, this is successful in over 90% of patients. Lower procedure-related complications and the relative non-invasive nature of endoscopic treatment has relegated surgical manement to a subsidiary role. Hospitalization time rarely exceeds 1-2 days. In most patients with advanced malignant disease and short life expectancy, stenting affords effective palliation. For the majority of patients endoscopic management is preferable to the percutaneous transhepatic approach due to lower overall mortality and morbidity. Stent occlusion necessitating replacement remains a problem, but improvements in this area can be expected. New plastic stent designs are undergoing investigation. Expandable metallic stents are promising but controlled comparative trials with conventional plastic prostheses are needed. Use of expandable stents should be judicious since these cannot be removed. In the future we can look forward to advances in peroral cholangioscopic technology which may permit targeted treatment of intraductal biliary malignancies.
Today, several endoscopic options for the management of benign or malignant obstruction of the biliary tract are available. Aside from biliary stone extraction, bile duct stenting is the most prevalent method employed. All other procedures are either infrequent or still in the experimental stage. Endoscopic decompression of the bile duct can be achieved with either external or internal drainage. For external drainage a naso-biliary catheter is introduced, whereas for internal drainage a stent (or endoprosthesis) is inserted. Stent placement has become the standard technique. Naso-biliary catheter placement is reserved mainly for bile duct irrigation in the treatment of bacterial cholangitis or during lithotripsy (extracorporeal shock wave or electrohydraulic). Endoscopic Drainage
The method of endoscopic stent placement is based on the Seldinger technique. The prosthesis is pushed coaxially over a flexible guide-wire into the bile duct by means of a pusher tube. A small papillotomy facilitates insertion. Since its introduction in 1979 [1], the endoscopic method has proven to be superior to Reprint requests: Nib Soehendra, M.D., Department of Endoscopic Surgery, University Hospital, Hamburg-Eppendorf, Martinistr. 52, 2000 Hamburg 20, Federal Republic of Germany.
the percutaneous transhepatic method. Comparing published series [2], endoscopic transpapillary drainage had lower rates of morbidity and 30-day mortality (Table 1). According to a prospective comparative study by Speer and coworkers [3], endoscopic stent placement was more successful than the percutaneous transhepatic procedure (89% versus 76%). Morbidity was 67% and 30-day mortality 33% in the transhepatically treated group, and 19% and 15%, respectively, for endoscopic treatment. These figures represent only results for plastic endoprosthesis placement. More recently, introduction of the self-expandable metallic stent has made the percutaneous transhepatic approach more appealing. These new stents can be inserted through a relatively thin catheter, reducing the risks of percutaneous transhepatic stenting [4, 5]. However, the endoscopic transpapillary approach is still less invasive, therefore the percutaneous transhepatic method should be reserved for endoscopic failures. The combined percutaneous transhepatic and endoscopic approach ("rendez-vous" procedure) enables the implantation of a plastic prosthesis of larger diameter with less trauma to the liver [6-8]. Expandable Metallic Stent The expandable metallic stent was pioneered by Charles Dotter in 1969 [9]. Since then, a number of different expandable stents have been developed for placement in blood vessels and the urethra [9]. Experience with expandable metallic stents in the biliary tract is primarily limited to the Wallstent and the Gianturco stent, both of which are self-expanding. The Gianturco stent (Cook, Bloomington, Indiana, U.S.A.) is made of stainless steel wire measuring 0.016 to 0.018 inch in diameter bent in a zig-zag pattern to form a cylinder. The stent is introduced in a compressed state through a 12-F teflon delivery catheter and, upon release, expands to a diameter of 12 ram. The length of each stent is 2.5-3.0 cm. The stent has been modified by adding monofilament nylon suture to connect the legs of the stent [11]. Connecting the stent legs serves to control stent expansion as it is released from the delivery catheter and to limit stent expansion to a desired diameter. In addition, stents can be adjoined in an end-to-end fashion, thereby increasing the length of the stent. In a European multicenter trial comprising i I patients with
N. Soehendra et al.: Endoscopic Therapy for Biliary Obstruction Table 1. Endoscopic
versus stenting: published series."
percutaneous
transhepatic
1067
biliary
Parameter
Endoscopic ( 1 0 )
Percutaneous transhepatic 03)
Total no. of pts. Successful insertion Successful drainage 30-day mortality Early complications
2,138 87% 85% 13% 19%
1,160 90% 69% 23% 28%
"As collected in [2]. ( ) Number of series.
benign and 16 patients with malignant biliary strictures, immediate relief of jaundice and cholangitis was achieved in all patients that underwent Gianturco stent placement by the percutaneous transhepatic approach. However, in a follow-up Period of 6-21 months, 2 patients with benign strictures developed cholangitis and 9 of 16 patients with malignant strictures developed recurrent jaundice. The primary cause of recurrent Jaundice was ingrowth of tumor between the stent wires; migration was observed in 2 patients [12]. In another study, recurrent jaundice was observed in 5 (22%) of 23 stented Patients during a follow-up period of 2-59 weeks [13]. The results for patients with postoperative biliary strictures seem to be better; in one study with a 6 month follow-up period, Occlusion of the Gianturco stent was not observed [14]. The Wallstent stent (Medinvent SA, Lausanne, Switzerland) is Woven in the form of a tubular mesh from stainless steel alloy filaments measuring 0.09 mm in diameter. The crosspoints are not soldered, giving the stent considerable pliability. The stent is loaded by the manufacturer onto a delivery catheter with a Constraining membrane to keep the stent compressed. Upon retraction of the constraining membrane, the stent is released and expands. The Wallstent stent is relatively easy to insert through a 7-F or 9-F delivery catheter and expands to a lumen of 7-10 mm. Due to the large inner diameter, this stent provides excellent initial drainage, and there is practically never blockage from Sludge, a common problem with plastic prostheses. However, despite the fine wire mesh, the problem of stent blockage by tumor extension or invasion remains. The rate of recurrent Jaundice was 18-42% in mean or median follow-up periods of 4--5 months [15-18]. By comparison, the rate of clogging for Plastic prostheses in malignant distal obstructions was around 20% for the same length of time [19, 20]. Prospective studies COmparing expandable metallic and plastic prostheses are awaited. Commercially available expandable metallic stents are costly and are not the final answer to the problem of late occlusion. Whether stent covering will resolve this problem remains to be Seen. In one experimental study evaluating silicone and polyether-polyurethane coated metallic stents in dogs, stents remained patent up to 24 weeks of follow-up, but varying degrees of luminal diameter reduction due to hyperplastic papillary naUcosal proliferation were noted in all animals [21]. Expandable stents, unlike plastic prostheses, cannot be removed or exchanged after placement, and since the long-term biotolerante of these permanent foreign bodies is not known, patients should be carefully selected for these new prostheses.
Fig. 1. Radiographs showing an endoscopicatly inserted 10-F "gutter" stent in a patient with malignant distal obstruction of the common bile duct. The distal lumen of the prosthesis is placed beyond the papilla with the "gutter" appendage extending into the duodenum.
Plastic Stent
The most widely employed plastic stents have an outer diameter of 10-F (3.3 mm). Stents of larger caliber are more difficult to insert. Plastic prostheses can be manufactured from various materials, including teflon, polyethylene, and polyurethane. All plastic prostheses are hampered by a high rate of blockage due to the accumulation of biliary sludge, probably encouraged by bacteria [22] and the reflux of chyme from the duodenum. Analysis of sludge extracted from occluded stents has demonstrated ample bacterial contamination (mostly E. coli) and plant fibers [23]. It is postulated that a layer of bacterial bio-film initially collects on the relatively rough surface of the plastic prosthesis which encourages adherence of sludge [24]. Antibacterial coating of stents has been proposed to reduce bacterial contamination. An in vitro study demonstrated that impregnation with silver or chlorhexidine significantly reduced E. coli growth [25]. However, this did not translate to prolonged stent patency in in vivo studies; stents coated with antibiotics such as cephalosporin had an occlusion rate no different than that of conventional stents [26]. In order to counteract post-stenting duodeno-biliary reflux that may introduce bacteria as well as chyme, we recently developed a stent which could be placed intraductally above the papilla, leaving only a transpapillary "gutter" appendage for the purpose of stent removal (Fig. 1). We compared three different types of prostheses in a clinical study [27] comprising patients with distal malignant biliary obstruction. Patients in group A received a conventional stent, in group B a silverimpregnated stent, and in group C the gutter stent. All stents were 10-F in outer diameter made of teflon formed in a "pigtail'" shape. Excluding patients with stent dislogement during the study, there was no difference in the duration of patency
Fig. 2. A. Endoscopic retrograde cholangiography showing an approximately 5 cm impacted stone in the proximal common bile duct. B. Successful fragmentation with cholangioscopic electrohydraulic lithotripsy. C. Complete clearance of the bile duct.
N. Soehendra et a|.: Endoscopic Therapy for Biliary Obstruction
1069
Fig. 3. A 10-year-oldgirl with a stricture of the proximal common bile duct following iatrogenic injury during cholecystectomy for cholecystolithiasis. A. Endoscopic retrograde cholangiography after operative biliary reconstruction showing a tight stricture at the proximal common bile duct. B. 12-F "pig-tail" prosthesis placed endoscopically for ductal dilatation. C. Status after 6 months of stent placement demonstrating satisfactory dilatation. between the conventional and silver-impregnated prostheses (111 vs 115 days, respectively), but longer patency for the gutter stem, averaging 20 days in excess of the conventional or silver prosthesis. As part of this study, we also analyzed the sites of stent Occlusion. Occlusions tended to be located along the lateral holes and the anchoring flaps. One explanation for this might be the relatively rough edges of the holes, which are more prone to trap sludge and food fibers. Experimental studies have shown that stent construction from teflon, which affords a smoother surface compared to other plastics, and the absence of side holes, are associated with a 6-fold lower incidence of sludge formation. Compared to prostheses with side holes, this reduction was statistically significant; biliary drainage was not comPromised by the absence of side holes [28]. In fact, side holes naay decrease bile flow due to flow turbulence [29]. Based on this experience, we only use prostheses without side holes. The flaps, which prevent dislocation, are constructed so that the lumen of the catheter is not opened. Comparative clinical trials are needed to show whether this type of prosthesis indeed increases stent longevity.
Benign Obstruction Stones are the most common cause of benign biliary obstruction. Endoscopic papillotomy followed by stone extraction is the treatment of choice. In patients with impacted or giant stones, sophisticated techniques such as electrohydraulic lithOtripsy using the mother-baby scope system can be successfully applied (Fig. 2). Benign strictures appear mainly in the distal or mid portion of the Common bile duct. Multiple strictures involving both intra-
and extrahepatic ducts, as can be seen in sclerosing cholangitis, are relatively rare and poorly suited for endoscopic treatment. Distal strictures extending several centimeters in length are usually due to chronic pancreatitis, whereas mid strictures are usually iatrogenic or due to postoperative scarring. Endoscopic dilatation of benign strictures can be performed using the balloon (Microvasive Inc., Watertown, Massachusetts, U.S.A.) or dilating catheter (Wilson-Cook Medical Inc., Winston-Salem, North Carolina, U.S.A.). A long-term (average of 24 months) follow-up showed dilatation to be effective. Thirty-five patients with major ductal strictures due to primary sclerosing cholangitis experienced significant improvement, as measured by hospitalization rates, laboratory data, and radiographic appearance. However, repeat dilatations at 6 months and yearly intervals thereafter were necessary to maintain improvement [30]. To achieve more permanent dilatation, placement of a large bore or several prostheses for several months is recommended. Approximately 80% of patients with 7-F to 10-F stents in place for at least 3 months attained satisfactory dilatation results (Fig. 3). Possible complications are rare and are mostly related to papillotomy. In published studies to date, there was no procedure-related mortality or morbidity, and stent occlusion was rarely observed [31-33]. In benign postoperative strictures, results were better in patients diagnosed and treated earlier; patients in whom strictures were treated within the first 3 months of surgery had response rates of 90% compared to 62% in patients with delayed treatment [33]. More recently, the expandable metallic stent has been recommended for management of benign strictures. Short-term results (follow-up to 21 months) appeared favorable, although sample size was small [12, 14]. However, as noted above, the
1070
World J. Surg. Vol. 16, No. 6, Nov./Dec. 1992
long-term biotolerance of these stents remains an area of concern. These stents may occlude (but cannot be removed) and long-term clinical trials comparing expandable stents with conventional plastic prostheses are needed before expandable stents can be generally recommended. One appropriate indication at this time might be postoperative anastomotic strictures not amendable to surgical revision.
Malignant Obstruction Endoscopic drainage is a palliative measure for malignant biliary obstruction. Obstruction may be due to primary tumors of the biliary tract (papilla, bile duct, gallbladder), cancer of the pancreatic head, or metastases. While endoscopic biliary drainage is technically feasible in most patients, it is important to carefully select patients most likely to benefit from this treatment. It is usually easy to stent obstructions of the common bile duct, but obstructions involving the bifurcation and intrahepatic branches can be more problematic. These proximal obstructions are due to cholangiocarcinoma, gallbladder carcinoma, or metastases. In patients with high grade biliary obstruction, contrast media injection under pressure with the aid of the balloon catheter is helpful to define the anatomy and extent of the stricture. This is particularly important in obstruction at the hepatic bifurcation, where tumor growth may extend into the intrahepatic ducts. Even in cases of seemingly complete obstruction, injection under pressure may " o p e n " the stricture. If bilateral ductal involvement is documented, both hepatic ducts should be stented to avoid complications of incomplete drainage (Fig. 4). Unilateral drainage has a 32-38% chance of resulting cholangitis, with sepsis the main cause of death. The 30-day mortality rate was reported to be 29% to 37% [34, 35]. Strictures approaching or involving the hepatic bifurcation are best managed initially with bilateral placement o f a 7-F stent since larger stents are technically more difficult to place. Use of the hydrophilic polymer-coated steerable guide wire (Terumo Corp., Tokyo, Japan) facilitates passage of tight or sharply angulated strictures. Seven to 14 days later, the stents can be exchanged for 10-F stents. Stent exchanges are facilitated by the stent retriever [36] (Wilson-Cook Medical Inc., WinstonSalem, North Carolina, U.S.A.). Unsuccessful endoscopic drainage after contrast media filling of an obstructed duct is an indication for immediate percutaneous transhepatic drainage. The endoscopic and percutaneous methods can be combined ("rendez-vous" procedure) in order to minimize liver trauma. The entire procedure should be completed in one session in order to prevent cholangitis [8]. Because of the high risk of cholangitis, antibiotic prophylaxis is strongly recommended. In addition, contrast media should be mixed with an antibiotic such as gentamycin. In our opinion, type III Klatskin tumors with involvement of branches of the left and/or right hepatic duct are not suited for endoscopic drainage. Insufficient drainage of contaminated bile inflates the risk of bacterial cholangitis. Endoscopic biliary drainage is primarily advocated for extrahepatic duct obstruction in high risk patients. For patients that are good surgical candidates, endoscopic palliation must be weighed against surgical bypass. The overall rate of procedure related complications of endoscopic stenting has been reported
Fig. 4. Klatskin type II tumor. Successful drainage of the right and left hepatic ducts with two 7-F prostheses.
to be 3-4%, with mortality
World Journal of Surgery © 1992 by the Soci~t~ lnternationale de Chirurgie
Endoscopic Therapy for Biliary Obstruction Nib Soehendra, M.D., K e n n e t h F. Binmoeller, M.D., and H o r s t Grimm, M.D. Department of Endoscopic Surgery, University Hospital of Hamburg, Hamburg, Federal Republic of Germany Endoscopic management of biliary obstruction is feasible in most patients and has emerged as standard treatment. Aside from the removal of bile duct stones, placement of a biliary stent is the most commonly employed modality of management. In experienced hands, this is successful in over 90% of patients. Lower procedure-related complications and the relative non-invasive nature of endoscopic treatment has relegated surgical manement to a subsidiary role. Hospitalization time rarely exceeds 1-2 days. In most patients with advanced malignant disease and short life expectancy, stenting affords effective palliation. For the majority of patients endoscopic management is preferable to the percutaneous transhepatic approach due to lower overall mortality and morbidity. Stent occlusion necessitating replacement remains a problem, but improvements in this area can be expected. New plastic stent designs are undergoing investigation. Expandable metallic stents are promising but controlled comparative trials with conventional plastic prostheses are needed. Use of expandable stents should be judicious since these cannot be removed. In the future we can look forward to advances in peroral cholangioscopic technology which may permit targeted treatment of intraductal biliary malignancies.
Today, several endoscopic options for the management of benign or malignant obstruction of the biliary tract are available. Aside from biliary stone extraction, bile duct stenting is the most prevalent method employed. All other procedures are either infrequent or still in the experimental stage. Endoscopic decompression of the bile duct can be achieved with either external or internal drainage. For external drainage a naso-biliary catheter is introduced, whereas for internal drainage a stent (or endoprosthesis) is inserted. Stent placement has become the standard technique. Naso-biliary catheter placement is reserved mainly for bile duct irrigation in the treatment of bacterial cholangitis or during lithotripsy (extracorporeal shock wave or electrohydraulic). Endoscopic Drainage
The method of endoscopic stent placement is based on the Seldinger technique. The prosthesis is pushed coaxially over a flexible guide-wire into the bile duct by means of a pusher tube. A small papillotomy facilitates insertion. Since its introduction in 1979 [1], the endoscopic method has proven to be superior to Reprint requests: Nib Soehendra, M.D., Department of Endoscopic Surgery, University Hospital, Hamburg-Eppendorf, Martinistr. 52, 2000 Hamburg 20, Federal Republic of Germany.
the percutaneous transhepatic method. Comparing published series [2], endoscopic transpapillary drainage had lower rates of morbidity and 30-day mortality (Table 1). According to a prospective comparative study by Speer and coworkers [3], endoscopic stent placement was more successful than the percutaneous transhepatic procedure (89% versus 76%). Morbidity was 67% and 30-day mortality 33% in the transhepatically treated group, and 19% and 15%, respectively, for endoscopic treatment. These figures represent only results for plastic endoprosthesis placement. More recently, introduction of the self-expandable metallic stent has made the percutaneous transhepatic approach more appealing. These new stents can be inserted through a relatively thin catheter, reducing the risks of percutaneous transhepatic stenting [4, 5]. However, the endoscopic transpapillary approach is still less invasive, therefore the percutaneous transhepatic method should be reserved for endoscopic failures. The combined percutaneous transhepatic and endoscopic approach ("rendez-vous" procedure) enables the implantation of a plastic prosthesis of larger diameter with less trauma to the liver [6-8]. Expandable Metallic Stent The expandable metallic stent was pioneered by Charles Dotter in 1969 [9]. Since then, a number of different expandable stents have been developed for placement in blood vessels and the urethra [9]. Experience with expandable metallic stents in the biliary tract is primarily limited to the Wallstent and the Gianturco stent, both of which are self-expanding. The Gianturco stent (Cook, Bloomington, Indiana, U.S.A.) is made of stainless steel wire measuring 0.016 to 0.018 inch in diameter bent in a zig-zag pattern to form a cylinder. The stent is introduced in a compressed state through a 12-F teflon delivery catheter and, upon release, expands to a diameter of 12 ram. The length of each stent is 2.5-3.0 cm. The stent has been modified by adding monofilament nylon suture to connect the legs of the stent [11]. Connecting the stent legs serves to control stent expansion as it is released from the delivery catheter and to limit stent expansion to a desired diameter. In addition, stents can be adjoined in an end-to-end fashion, thereby increasing the length of the stent. In a European multicenter trial comprising i I patients with
N. Soehendra et al.: Endoscopic Therapy for Biliary Obstruction Table 1. Endoscopic
versus stenting: published series."
percutaneous
transhepatic
1067
biliary
Parameter
Endoscopic ( 1 0 )
Percutaneous transhepatic 03)
Total no. of pts. Successful insertion Successful drainage 30-day mortality Early complications
2,138 87% 85% 13% 19%
1,160 90% 69% 23% 28%
"As collected in [2]. ( ) Number of series.
benign and 16 patients with malignant biliary strictures, immediate relief of jaundice and cholangitis was achieved in all patients that underwent Gianturco stent placement by the percutaneous transhepatic approach. However, in a follow-up Period of 6-21 months, 2 patients with benign strictures developed cholangitis and 9 of 16 patients with malignant strictures developed recurrent jaundice. The primary cause of recurrent Jaundice was ingrowth of tumor between the stent wires; migration was observed in 2 patients [12]. In another study, recurrent jaundice was observed in 5 (22%) of 23 stented Patients during a follow-up period of 2-59 weeks [13]. The results for patients with postoperative biliary strictures seem to be better; in one study with a 6 month follow-up period, Occlusion of the Gianturco stent was not observed [14]. The Wallstent stent (Medinvent SA, Lausanne, Switzerland) is Woven in the form of a tubular mesh from stainless steel alloy filaments measuring 0.09 mm in diameter. The crosspoints are not soldered, giving the stent considerable pliability. The stent is loaded by the manufacturer onto a delivery catheter with a Constraining membrane to keep the stent compressed. Upon retraction of the constraining membrane, the stent is released and expands. The Wallstent stent is relatively easy to insert through a 7-F or 9-F delivery catheter and expands to a lumen of 7-10 mm. Due to the large inner diameter, this stent provides excellent initial drainage, and there is practically never blockage from Sludge, a common problem with plastic prostheses. However, despite the fine wire mesh, the problem of stent blockage by tumor extension or invasion remains. The rate of recurrent Jaundice was 18-42% in mean or median follow-up periods of 4--5 months [15-18]. By comparison, the rate of clogging for Plastic prostheses in malignant distal obstructions was around 20% for the same length of time [19, 20]. Prospective studies COmparing expandable metallic and plastic prostheses are awaited. Commercially available expandable metallic stents are costly and are not the final answer to the problem of late occlusion. Whether stent covering will resolve this problem remains to be Seen. In one experimental study evaluating silicone and polyether-polyurethane coated metallic stents in dogs, stents remained patent up to 24 weeks of follow-up, but varying degrees of luminal diameter reduction due to hyperplastic papillary naUcosal proliferation were noted in all animals [21]. Expandable stents, unlike plastic prostheses, cannot be removed or exchanged after placement, and since the long-term biotolerante of these permanent foreign bodies is not known, patients should be carefully selected for these new prostheses.
Fig. 1. Radiographs showing an endoscopicatly inserted 10-F "gutter" stent in a patient with malignant distal obstruction of the common bile duct. The distal lumen of the prosthesis is placed beyond the papilla with the "gutter" appendage extending into the duodenum.
Plastic Stent
The most widely employed plastic stents have an outer diameter of 10-F (3.3 mm). Stents of larger caliber are more difficult to insert. Plastic prostheses can be manufactured from various materials, including teflon, polyethylene, and polyurethane. All plastic prostheses are hampered by a high rate of blockage due to the accumulation of biliary sludge, probably encouraged by bacteria [22] and the reflux of chyme from the duodenum. Analysis of sludge extracted from occluded stents has demonstrated ample bacterial contamination (mostly E. coli) and plant fibers [23]. It is postulated that a layer of bacterial bio-film initially collects on the relatively rough surface of the plastic prosthesis which encourages adherence of sludge [24]. Antibacterial coating of stents has been proposed to reduce bacterial contamination. An in vitro study demonstrated that impregnation with silver or chlorhexidine significantly reduced E. coli growth [25]. However, this did not translate to prolonged stent patency in in vivo studies; stents coated with antibiotics such as cephalosporin had an occlusion rate no different than that of conventional stents [26]. In order to counteract post-stenting duodeno-biliary reflux that may introduce bacteria as well as chyme, we recently developed a stent which could be placed intraductally above the papilla, leaving only a transpapillary "gutter" appendage for the purpose of stent removal (Fig. 1). We compared three different types of prostheses in a clinical study [27] comprising patients with distal malignant biliary obstruction. Patients in group A received a conventional stent, in group B a silverimpregnated stent, and in group C the gutter stent. All stents were 10-F in outer diameter made of teflon formed in a "pigtail'" shape. Excluding patients with stent dislogement during the study, there was no difference in the duration of patency
Fig. 2. A. Endoscopic retrograde cholangiography showing an approximately 5 cm impacted stone in the proximal common bile duct. B. Successful fragmentation with cholangioscopic electrohydraulic lithotripsy. C. Complete clearance of the bile duct.
N. Soehendra et a|.: Endoscopic Therapy for Biliary Obstruction
1069
Fig. 3. A 10-year-oldgirl with a stricture of the proximal common bile duct following iatrogenic injury during cholecystectomy for cholecystolithiasis. A. Endoscopic retrograde cholangiography after operative biliary reconstruction showing a tight stricture at the proximal common bile duct. B. 12-F "pig-tail" prosthesis placed endoscopically for ductal dilatation. C. Status after 6 months of stent placement demonstrating satisfactory dilatation. between the conventional and silver-impregnated prostheses (111 vs 115 days, respectively), but longer patency for the gutter stem, averaging 20 days in excess of the conventional or silver prosthesis. As part of this study, we also analyzed the sites of stent Occlusion. Occlusions tended to be located along the lateral holes and the anchoring flaps. One explanation for this might be the relatively rough edges of the holes, which are more prone to trap sludge and food fibers. Experimental studies have shown that stent construction from teflon, which affords a smoother surface compared to other plastics, and the absence of side holes, are associated with a 6-fold lower incidence of sludge formation. Compared to prostheses with side holes, this reduction was statistically significant; biliary drainage was not comPromised by the absence of side holes [28]. In fact, side holes naay decrease bile flow due to flow turbulence [29]. Based on this experience, we only use prostheses without side holes. The flaps, which prevent dislocation, are constructed so that the lumen of the catheter is not opened. Comparative clinical trials are needed to show whether this type of prosthesis indeed increases stent longevity.
Benign Obstruction Stones are the most common cause of benign biliary obstruction. Endoscopic papillotomy followed by stone extraction is the treatment of choice. In patients with impacted or giant stones, sophisticated techniques such as electrohydraulic lithOtripsy using the mother-baby scope system can be successfully applied (Fig. 2). Benign strictures appear mainly in the distal or mid portion of the Common bile duct. Multiple strictures involving both intra-
and extrahepatic ducts, as can be seen in sclerosing cholangitis, are relatively rare and poorly suited for endoscopic treatment. Distal strictures extending several centimeters in length are usually due to chronic pancreatitis, whereas mid strictures are usually iatrogenic or due to postoperative scarring. Endoscopic dilatation of benign strictures can be performed using the balloon (Microvasive Inc., Watertown, Massachusetts, U.S.A.) or dilating catheter (Wilson-Cook Medical Inc., Winston-Salem, North Carolina, U.S.A.). A long-term (average of 24 months) follow-up showed dilatation to be effective. Thirty-five patients with major ductal strictures due to primary sclerosing cholangitis experienced significant improvement, as measured by hospitalization rates, laboratory data, and radiographic appearance. However, repeat dilatations at 6 months and yearly intervals thereafter were necessary to maintain improvement [30]. To achieve more permanent dilatation, placement of a large bore or several prostheses for several months is recommended. Approximately 80% of patients with 7-F to 10-F stents in place for at least 3 months attained satisfactory dilatation results (Fig. 3). Possible complications are rare and are mostly related to papillotomy. In published studies to date, there was no procedure-related mortality or morbidity, and stent occlusion was rarely observed [31-33]. In benign postoperative strictures, results were better in patients diagnosed and treated earlier; patients in whom strictures were treated within the first 3 months of surgery had response rates of 90% compared to 62% in patients with delayed treatment [33]. More recently, the expandable metallic stent has been recommended for management of benign strictures. Short-term results (follow-up to 21 months) appeared favorable, although sample size was small [12, 14]. However, as noted above, the
1070
World J. Surg. Vol. 16, No. 6, Nov./Dec. 1992
long-term biotolerance of these stents remains an area of concern. These stents may occlude (but cannot be removed) and long-term clinical trials comparing expandable stents with conventional plastic prostheses are needed before expandable stents can be generally recommended. One appropriate indication at this time might be postoperative anastomotic strictures not amendable to surgical revision.
Malignant Obstruction Endoscopic drainage is a palliative measure for malignant biliary obstruction. Obstruction may be due to primary tumors of the biliary tract (papilla, bile duct, gallbladder), cancer of the pancreatic head, or metastases. While endoscopic biliary drainage is technically feasible in most patients, it is important to carefully select patients most likely to benefit from this treatment. It is usually easy to stent obstructions of the common bile duct, but obstructions involving the bifurcation and intrahepatic branches can be more problematic. These proximal obstructions are due to cholangiocarcinoma, gallbladder carcinoma, or metastases. In patients with high grade biliary obstruction, contrast media injection under pressure with the aid of the balloon catheter is helpful to define the anatomy and extent of the stricture. This is particularly important in obstruction at the hepatic bifurcation, where tumor growth may extend into the intrahepatic ducts. Even in cases of seemingly complete obstruction, injection under pressure may " o p e n " the stricture. If bilateral ductal involvement is documented, both hepatic ducts should be stented to avoid complications of incomplete drainage (Fig. 4). Unilateral drainage has a 32-38% chance of resulting cholangitis, with sepsis the main cause of death. The 30-day mortality rate was reported to be 29% to 37% [34, 35]. Strictures approaching or involving the hepatic bifurcation are best managed initially with bilateral placement o f a 7-F stent since larger stents are technically more difficult to place. Use of the hydrophilic polymer-coated steerable guide wire (Terumo Corp., Tokyo, Japan) facilitates passage of tight or sharply angulated strictures. Seven to 14 days later, the stents can be exchanged for 10-F stents. Stent exchanges are facilitated by the stent retriever [36] (Wilson-Cook Medical Inc., WinstonSalem, North Carolina, U.S.A.). Unsuccessful endoscopic drainage after contrast media filling of an obstructed duct is an indication for immediate percutaneous transhepatic drainage. The endoscopic and percutaneous methods can be combined ("rendez-vous" procedure) in order to minimize liver trauma. The entire procedure should be completed in one session in order to prevent cholangitis [8]. Because of the high risk of cholangitis, antibiotic prophylaxis is strongly recommended. In addition, contrast media should be mixed with an antibiotic such as gentamycin. In our opinion, type III Klatskin tumors with involvement of branches of the left and/or right hepatic duct are not suited for endoscopic drainage. Insufficient drainage of contaminated bile inflates the risk of bacterial cholangitis. Endoscopic biliary drainage is primarily advocated for extrahepatic duct obstruction in high risk patients. For patients that are good surgical candidates, endoscopic palliation must be weighed against surgical bypass. The overall rate of procedure related complications of endoscopic stenting has been reported
Fig. 4. Klatskin type II tumor. Successful drainage of the right and left hepatic ducts with two 7-F prostheses.
to be 3-4%, with mortality
