Catheter-Free Survival After Primary Percutaneous Stenting of Malignant Bile Duct Obstruction

Vascular and Interventional Radiology Original Research Thornton et al. Catheter-Free Survival After Bile Duct Stenting Vascular and Interventional Radiology Original Research Raymond H. Thornton 1 Benjamin S. Frank 2 Anne M. Covey 1 Majid Maybody 1 Stephen B. Solomon 1 George I. Getrajdman 1 Karen T. Brown 1 Thornton RH, Frank B, Covey AM, et al. Keywords: catheter free, malignant bile duct obstruction, percutaneous stenting DOI:10.2214/AJR.10.6069 Received November 5, 2010; accepted after revision January 13, 2011. S. B. Solomon is a consultant to and has received research grants from AngioDynamics and GE Healthcare. He is also a consultant to Johnson & Johnson and Medtronic and has received a research grant from Visualase. 1 Interventional Radiology Service, Department of Radiology, Memorial Sloan Kettering Cancer Center, 1275 York Ave, H-118, New York, NY 10065. Address correspondence to R. H. Thornton (Thorntor@mskcc.org). 2 Weill Cornell Medical College, New York, NY. WEB This is a Web exclusive article. AJR 2011; 197:W514 W518 0361 803X/11/1973 W514 American Roentgen Ray Society Catheter-Free Survival After Primary Percutaneous Stenting of Malignant Bile Duct Obstruction OBJECTIVE. The purpose of this article is to review the outcomes of patients with malignant biliary obstruction treated by immediate placement of a metallic stent at the time of percutaneous biliary drainage. MATERIALS AND METHODS. By a search of our PACS, we retrospectively identified all patients who underwent biliary stent placement at our interventional radiology section between January 2003 and December 2008. Of 564 biliary stent procedures, 71 primary percutaneous stents were placed at the time of biliary drainage; the other 493 were secondary percutaneous stents, placed at various durations after biliary drainage. Medical records and procedure reports were reviewed for the primary percutaneous stent group recording time to removal of the percutaneous catheter, periprocedural complications, subsequent biliary interventions, overall survival, and catheter-free survival. RESULTS. After placement of primary percutaneous stents, 97% of patients were able to have their transhepatic catheter removed, and in 73% of patients, this 5-French catheter was removed within 24 hours. Major complications after placement of primary percutaneous stents included bile peritonitis requiring IV analgesia for less than 48 hours (5.6%) and fever with leukocytosis treated with IV antibiotics (8.5%) for 48 72 hours. Median overall survival was 165 days, and 90% of patients had 100% catheter-free survival. Overall, 94% of survival days were catheter free. CONCLUSION. Placement of primary percutaneous stents when clinically appropriate has the advantage of avoiding an externalized drainage catheter and its attendant lifestyle limitations and complications. Most patients with malignant biliary obstruction, for whom expected survival is short, will experience 100% catheter-free survival after placement of a primary percutaneous stent. M alignant biliary obstruction often signals incurable disease with survival measured in months [1, 2]. An ideal treatment of malignant biliary obstruction should relieve biliary obstruction in as minimally invasive a way as possible and have durable patency for the remainder of the patient s life. Endoscopic placement of biliary stents comes closest to achieving these goals. In some patients, however, endoscopic drainage is not possible and percutaneous biliary drainage is required. This is often accomplished by placement of a percutaneous biliary drainage catheter, which may later be converted to an internal metallic stent. This two-step approach, or secondary percutaneous stent placement, requires the patient to have a percutaneous catheter for a period of time between the biliary drainage and stent placement. In a recent report on secondary percutaneous stent placement, it was characteristic for patients with cancer to live with percutaneous drains for 2 6 weeks before internalization with a metal stent [3]. Percutaneous biliary catheters are associated with lifestyle limitations and potential complications, including insertion-site pain, dislodgement, catheter dysfunction, and pericatheter leakage of bile or ascites. To manage these issues, unscheduled catheter care visits as frequent as once a month are sometimes required [4]. Often urgent, such inconveniences are suboptimal for any patient but are particularly onerous when they interrupt at-home palliative care. Primary percutaneous stent placement entails placement of a metallic stent at the time of biliary drainage in a one-step procedure. The percutaneous catheter can then be removed as soon as the stent is shown to be W514 AJR:197, September 2011

Catheter-Free Survival After Bile Duct Stenting patent, freeing the patient from the need for a percutaneous drain and its attendant potential problems. The purpose of this study was to review the outcomes of patients with malignant biliary obstruction treated by primary percutaneous stent placement at our institution. Materials and Methods A waiver of consent for retrospective review was obtained from our institutional review board. A query of our PACS for procedures coded biliary stent from January 2003 to December 2008 was performed. Of 564 biliary stents placed during this time period, 71 were primary percutaneous stents and 493 were secondary percutaneous stents. We reviewed the medical records and imaging of patients who underwent primary percutaneous stent placement. All patients who underwent primary percutaneous stent placement had bile duct obstruction in the setting of known malignancy. Patient characteristics are recorded in Table 1. Description of Procedure After IV administration of ticarcillin and clavulanic acid for antibiotic prophylaxis, biliary drainage is performed under moderate sedation using IV midazolam and either meperidine or fentanyl. Once the biliary tree is accessed with a percutaneous access set (Neff, Cook Medical) from a suitable peripheral puncture, diagnostic cholangiography is performed. The biliary obstruction is then crossed and contrast agent is injected distal to the obstruction to reveal the length of biliary occlusion. It is not our practice to predilate the biliary stricture. The catheter is advanced into the small intestine, and contrast agent injection is performed to verify luminal patency and peristalsis. Through a 9-French sidearm vascular sheath and over wire, an appropriate length (8 or 10 mm) biliary stent (Wallstent, Boston Scientific) is advanced across the obstruction and deployed under fluoroscopic visualization. The immediate result is assessed by sheath cholangiography. If contrast material does not flow through the stent, additional coaxial stents may be placed or the stent may be balloon dilated, according to operator preference. Once sheath cholangiography reveals passage of contrast agent through the stent, a 5-French 40- cm Berenstein catheter (Soft-Vu, Angiodynamics) is inserted to preserve transhepatic biliary access, and this catheter is capped. The patient is admitted for overnight observation and returns for stent check the next morning. The stent is checked by retracting the Berenstein catheter to the cranial aspect of the stent and injecting contrast material. If stent geometry and flow through the stent into the intestine are deemed adequate, the catheter is removed, and the patient is catheter free. TABLE 1: Demographic Information of 71 Patients Variable Value Sex Female 36 (51) Male 35 (49) Age (y), median (range) 65 (38 89) Diagnosis Pancreas cancer 31 (44) Cholangiocarcinoma 17 (24) Colorectal carcinoma 8 (11) Gastric carcinoma 5 (7) Uterine carcinoma 3 (4) Gastrointestinal stromal tumor 2 (3) Breast cancer 1 (1) Esophageal cancer 1 (1) Duodenal carcinoma 1 (1) Appendiceal carcinoma 1 (1) Unknown primary carcinoma 1 (1) Prior hepatic resection No 68 (96) Yes 3 (4) a Bilioenteric anastomosis No 67 (94) Yes 4 (6) b Indwelling duodenal stent No 64 (90) Yes 7 (10) Note Except where noted otherwise, data are no. (%) of patients. Not all percentages in each category add up to 100% because of rounding. a Two patients had undergone right hepatectomy, and another had undergone right trisegmentectomy. b Four patients had undergone hepaticojejunostomy. If either stent geometry or flow is suboptimal, the stent may be dilated with a balloon or a coaxial stent placed, as required. Patients undergoing stent revision typically have either a successful repeat stent check with catheter removal the next morning or are discharged with an external biliary drain for subsequent removal as an outpatient. Procedural information for the study group is recorded in Table 2. Definitions We defined procedural success as the ability to remove the percutaneous catheter, rendering the patient catheter free. Catheter-free survival is defined as the number of days the patient survived without a percutaneous catheter after its removal, and percentage of catheter-free survival is catheter-free survival divided by total survival days. Results During the 5 years studied, 564 biliary stent procedures were performed. Of these, 71 (12.6%) were primary percutaneous stents and 493 (87.4%) were secondary percutaneous stents. In the primary percutaneous stents group, procedural success was achieved in all but two patients (97%). Removal of the Percutaneous Catheter In 52 patients with primary percutaneous stents (73%), the percutaneous biliary catheter was removed within 24 hours. This group included three patients whose percutaneous biliary access was removed immediately after stent deployment, with absorbable gelatin sponge (Gelfoam, Pfizer) embolization of the transhepatic tract. Sixty patients (85%) underwent removal of the percutaneous catheter within 48 hours. This group included eight additional patients; five patients required stent revision (addition of a coaxial stent) during the first stent check and underwent successful stent check with catheter removal the following day, and two patients had poor flow through unstented papillae, presumably due to papillary spasm induced by the presence of the catheter across the sphincter of Oddi. These two patients were treated by repositioning the 5-French catheter above the ampulla, and both underwent successful stent check with catheter removal the next day. The other patient underwent brachytherapy after primary percutaneous stent placement and underwent removal of the brachytherapy catheters within 48 hours. An additional nine patients underwent removal of their percutaneous catheter 3 27 days after primary percutaneous stent placement. Eight had suboptimal flow on stent check despite patent well-positioned stents. These patients had placement of 8-French external biliary drains that were subsequently capped; all were removed 3 27 days later. The other patient had bleeding around the 5-French catheter immediately after primary percutaneous stent placement. This bleeding was treated by placement of a 10-French external biliary drain to tamponade the tract. Bleeding immediately ceased, and the catheter was uneventfully removed 12 days later in an outpatient procedure. The procedural failures included two patients who required long-term percutaneous biliary drainage. In both cases, flow through AJR:197, September 2011 W515

Thornton et al. well-positioned patent biliary stents was inadequate because of duodenal disease. Complications Three types of complications were recognized within 30 days: bile leakage, infection, and bleeding. There were 10 major complications [5], including four patients (5.6%) who developed pain consistent with bile peritonitis. All four were treated with IV narcotics and symptoms resolved within 48 hours. Six patients (8.5%) developed fever and leukocytosis after the procedure, prompting IV antibiotic treatment of 48 72 hours. Minor complications included a skin site infection in one patient that responded to oral antibiotics. Pericatheter bleeding prompted placement of a 10-French external biliary drain and delayed catheter removal, as noted in the previous section. There was no procedure-related mortality. Repeat Intervention Eleven patients (15%) required a repeat intervention to reestablish stent patency during their survival period. Six had recurrent jaundice, and five had cholangitis. Median time to reintervention was 85 days (range, 28 490 days). Among these patients, four underwent definitive endoscopic stent revision and seven patients (10%) required repeat percutaneous drainage. Survival Median overall survival by the Kaplan- Meier method was 165 days (range, 3 1430 days), with 11 patients alive at the time of analysis (Fig. 1). Of the 69 patients whose percutaneous catheter was successfully removed after stent placement, 62 (90%) had 100% catheterfree survival. Among the seven patients who required percutaneous redrainage, 58% of their survival days were catheter free. Overall, 94% of all survival days were catheter free after primary percutaneous stents. Discussion Quality of life and optimal palliation are essential considerations when treating patients with malignant bile duct obstruction. Placement of percutaneous biliary drains, although effective for relief of jaundice and pruritus and for treatment of cholangitis, commits the patient to externalized catheters that can be difficult to manage and are prone to dislodge, occlude, and leak bile or ascites. Endoscopic placement of metallic biliary stents avoids many of these complications. Unfortunately, not every patient with bile duct obstruction can be managed endoscopically. TABLE 2: Primary Percutaneous Stent Placement Procedural Information No. (%) of Patients Variable (n = 71) Indication for drainage Jaundice 30 (42) Require lower bilirubin for chemotherapy 22 (31) Pruritus 11 (15) Cholangitis a 3 (4) Multiple indications b 5 (7) Level of biliary obstruction Common bile duct 40 (56) Common hepatic duct (Bismuth type I) 21 (30) Obstruction involving primary biliary confluence (Bismuth type II) 6 (8) Obstruction extending into secondary biliary confluences (Bismuth type III) 4 (6) Stent diameter (mm) 8 28 (39) 10 43 (61) No. of stents 1 61 (86) 2 (coaxial) 10 (14) Balloon dilation of stent No 55 (77) Yes c 16 (23) Note Not all percentages in each category add up to 100% because of rounding. a Patients with cholangitis, clinically defined as leukocytosis and fever in the setting of biliary obstruction with no other fever source, underwent primary percutaneous stent placement after 24 48 hours of IV antibiotic therapy. b Four patients required lower bilirubin for chemotherapy and also had pruritus; one patient had pruritus and fever. c Balloon diameters ranged from 4 to 10 mm. Secondary percutaneous stent placement begins with placement of a percutaneous biliary drain. After a period of time, the drainage catheter is exchanged for a biliary stent, and the percutaneous catheter can be removed. The timing of such an exchange is variable and typically practice specific. At our institution, for example, we prefer to wait 2 3 weeks to allow maturation of an epithelialized tract from the biliary tree to the skin, to minimize the risk of intraperitoneal bile leakage upon catheter removal. Clinical factors also play a role in the timing of the exchange. When the indication for drainage is to decrease the bilirubin to permit administration of chemotherapy, for example, some may wait for a clinically acceptable bilirubin nadir before abandoning percutaneous access and internalizing drainage, especially in cases of Bismuth type II IV biliary obstruction when the biliary tree is incompletely drained. In such cases of incomplete biliary drainage, cholangitis may develop in these contaminated but poorly drained portions of the biliary tree [6] in the interim. Because of the high rate of catheter-related complications, we have shifted our paradigm for percutaneous biliary drainage to favor primary percutaneous stents when possible. There are several reasons to expect that this change might be beneficial to the patient. The mean patency of biliary stents for malignant biliary obstruction is approximately 8.5 months [7], and in this study of 71 patients with malignant biliary obstruction resulting from a variety of primary cancers, the median overall survival was less than 6 months. Placement of a biliary stent at the time of drainage, therefore, has the potential to provide 100% catheter-free survival. Even if a patient outlives the patency of the stent and requires percutaneous reintervention, the survival interval preceding loss of stent patency is catheter free. In cases of high bile duct obstruction, primary percutaneous stents may offer the potential to avoid contamination of poorly drained portions of the biliary tree. Each time a biliary catheter is flushed to maintain patency, there is an opportunity to introduce bacteria into incompletely W516 AJR:197, September 2011

Catheter-Free Survival After Bile Duct Stenting Proportion Surviving (%) 100 75 50 25 0 0 drained liver segments, leading to cholangitis. Primary percutaneous stents eliminate this mode of potential superinfection. The technique of primary percutaneous stent placement is not new, having been described by several authors over the past two decades. Among these reports, however, the length of time an externalized catheter was maintained after primary percutaneous stent placement has varied considerably. Fifty-six patients with primary stents were included in a series by Lammer et al. [8] in 1990. They described placement of compressed gelatin sponge into the transhepatic tract, facilitating immediate catheter removal in uncomplicated cases, a technique subsequently also reported by McKeown et al. [9]. For cases with hemobilia, a flushing catheter was maintained for 2 days. In 1992, Lee et al. [10] described a group of 34 patients treated for malignant bile duct obstruction with metallic biliary endoprostheses. In that report, 13 patients with confirmed malignancy at the time of drainage received primary biliary stents. Eight-French external catheters were removed after 4 days on average (range, 1 10 days), after successful stent checks. An Italian group reported treatment of 50 patients with a single-step technique in 1994, where external catheters were maintained for 1 2 days [11]. In 1995, Hannesson et al. [12] described 55 patients with malignant biliary obstruction, of whom 18 were treated with primary stents, but the length of external drainage was not specified. 500 1000 1500 Time (d) From Stent Placement Fig. 1 Kaplan-Meier estimate of overall survival for patients with malignant biliary obstruction treated by placement of metallic stent at time of biliary drainage. In 2003, Inal et al. [13] published results comparing one- and two-stage biliary stent insertions with and without predilation of the biliary stricture. Among the 44 patients who received primary percutaneous stents in that study, no drainage catheter was inserted in 11 (25%). When immediate stent expansion appeared inadequate, the other patients had placement of a 6- to 8-French catheter to provide external biliary drainage for 1 13 days (mean, 3 days) until catheter cholangiograms showed adequate stent expansion and patency. There was no significant difference in mean stent patency or patient survival when biliary stenting was performed in one or two stages, with or without predilation. Twostage procedures and predilation, however, increased costs and early complication rates. In 2003, Yoshida et al. [14] redescribed the technique of one-step biliary stent placement in a patient with malignant biliary obstruction. A 9-French external catheter was left in place for 3 days. The same authors updated their experience with primary biliary stent placement in 2006 [15], reporting 14 cases of primary biliary stent placement, with 9-French catheters left in place for 2 3 days. For catheter removal, this group required both a normal stent check and biochemical evidence of resolving cholestasis. In 2008, results from five patients treated with one-step bile duct stents were reported [16]. A 6-French external catheter was placed overnight, clamped the next day for 6 24 hours, and removed when clinical findings confirmed adequate drainage. No time to catheter removal was specified. In their 2010 report on early infectious complications after biliary stent placement in 215 patients, Li Sol et al. [17] included four patients who had been treated with primary biliary stents. Percutaneous drains were removed 3 7 days after stent placement, when both stent check showed patency and serum bilirubin had decreased by more than 30%. This review adds to the literature our experience with 71 patients treated by primary percutaneous stent placement. In our experience, nearly three-quarters of patients had removal of the percutaneous catheter within 24 hours. This is not a trivial point. Typically, percutaneous biliary drainage requires an overnight hospital admission for observation. We have shown that the catheter may be removed safely in 73% of patients within 24 hours, or during the same overnight hospital stay, freeing the patient from a return visit and another procedure 2 7 days later. Also different from many prior reports, the 5-French percutaneous catheter that we place is intended as a transhepatic placeholder only and is not used as an external drain during the overnight observation period. We think that placement of the smaller diameter catheter minimizes the risk of bile leakage after its removal by permitting immediate recoil of the transhepatic tract after the brief dwell time of the larger procedural sheath. Only a few patients in our study group had symptoms of bile peritonitis after catheter removal. Larger 8- or 10-French drainage catheters were placed only when external biliary drainage was required because of inadequate antegrade passage of contrast agent through a newly inserted metallic stent. When 8- or 10-French drainage catheters were placed for drainage, prolonging the period of transhepatic dilation, we waited several weeks for tract maturation before catheter removal. During the study period, we did not advocate removal of percutaneous access immediately after primary percutaneous stent placement routinely, because a few patients will require short-term stent revision or continued external drainage. If access were abandoned immediately after primary percutaneous stent placement, this would have mandated either an endoscopic or a new biliary drainage procedure. In select instances, as for patients with ascites, immediate removal of the biliary access may be desirable and can be facilitated by absorbable gelatin sponge embolization of the transhepatic tract. The limitations of our study are typical of those of retrospective reviews. For exam- AJR:197, September 2011 W517

Thornton et al. ple, we have identified only those patients in whom a primary percutaneous stent was technically achieved. It was not possible to identify retrospectively which patients, if any, underwent percutaneous biliary drainage rather than primary percutaneous stent placement because of an intraprocedural event, such as hemobilia related to the drainage procedure or the inability to cross the obstruction at the time of initial biliary drainage. Among the 69 patients (97%) who could be freed from an externalized catheter, most had their catheter removed within 24 hours and 90% remained catheter free for their entire survival period after primary percutaneous stent placement. Overall, 94% of survival days were catheter free, and median survival was less than 6 months. The most common complications were fever with leukocytosis (8.5%) requiring IV antibiotics and bile peritonitis (5.6%) requiring IV analgesia. When patients with malignant biliary obstruction cannot be managed endoscopically, primary percutaneous stent placement is an important palliative option. References 1. Smith AC, Dowsett JF, Russell RC, Hatfield AR, Cotton PB. Randomised trial of endoscopic stenting versus surgical bypass in malignant low bileduct obstruction. Lancet 1994; 344:1655 1660 2. Robson PC, Heffernan N, Gonen M, et al. Prospective study of outcomes after percutaneous biliary drainage for malignant biliary obstruction. Ann Surg Oncol 2010; 17:2303 2311 3. Dahlstrand U, Sandblom G, Eriksson LG, Nyman R, Rasmussen IC. Primary patency of percutaneously inserted self-expanding metallic stents in patients with malignant biliary obstruction. HPB (Oxford) 2009; 11:358 363 4. Ulrich R, Thornton R, Brody LA, et al. Outcomes of patients undergoing percutaneous biliary drainage to reduce bilirubin for administration of chemotherapy. (abstr 340) Society of Interventional Radiology 35th Annual Scientific Meeting. Fairfax, VA: Society of Interventional Radiology, 2010 5. Sacks D, McClenny TE, Cardella JF, Lewis CA. Society of Interventional Radiology clinical practice guidelines. J Vasc Interv Radiol 2003; 14: S199 S202 6. Brown KT. Intrahepatic and extrahepatic biliary cancer (interventional radiologic techniques in hilar and intrahepatic biliary strictures). In: Blumgart L, ed. Surgery of the liver, biliary tract and pancreas. Philadelphia, PA: Elsevier, 2007: 814 822 7. Maybody M, Brown KT, Brody LA, et al. Primary patency of Wallstents in malignant bile duct obstruction: single vs. two or more noncoaxial stents. Cardiovasc Intervent Radiol 2009; 32:707 713 8. Lammer J, Klein GE, Kleinert R, Hausegger K, Einspieler R. Obstructive jaundice: use of expandable metal endoprosthesis for biliary drainage: work in progress. Radiology 1990; 177:789 792 9. McKeown BJ, Wong WL, Adam A. True singlestage percutaneous insertion of Wallstent biliary endoprosthesis. Clin Radiol 1993; 3:257 260 10. Lee MJ, Dawson SL, Mueller PR, Krebs TL, Saini S, Hahn PF. Palliation of malignant bile duct obstruction with metallic biliary endoprostheses: technique, results, and complications. J Vasc Interv Radiol 1992; 3:665 671 11. Ferro C, Perona F, Barile A, et al. Treatment of malignant occlusions of bile ducts by one-step technique using metallic endoprostheses: a multicenter study of 50 cases (in Italian). Radiol Med 1994; 88:74 78 12. Hannesson PH, Andrén-Sandberg A, Stridbeck H, Cwikiel W. Treatment of malignant biliary obstruction by percutaneous transhepatic insertion of expandable metallic stents. Eur Radiol 1995; 5:1 5 13. Inal M, Aksungur E, Akgul E, Oguz M, Seydaoglu G. Percutaneous placement of metallic stents in malignant biliary obstruction: one-stage or twostage procedure? Pre-dilate or not? Cardiovasc Intervent Radiol 2003; 26:40 45 14. Yoshida H, Tajiri T, Mamada Y, et al. One-step insertion of an expandable metallic stent for unresectable common bile duct carcinoma. J Nippon Med Sch 2003; 70:179 182 15. Yoshida H, Mamada Y, Taniai N, et al. One-step palliative treatment method for obstructive jaundice caused by unresectable malignancies by percutaneous transhepatic insertion of an expandable metallic stent. World J Gastroenterol 2006; 12: 2423 2426 16. Dambrauskas Z, Paskauskas S, Lizdenis P, et al. Percutaneous transhepatic biliary stenting: the first experience and results of the Hospital of Kaunas University of Medicine. Medicina (Kaunas) 2008; 44:969 976 17. Li Sol Y, Kim CW, Jeon UB, et al. Early infectious complications of percutaneous metallic stent insertion for malignant biliary obstruction. AJR 2010; 194:261 265 W518 AJR:197, September 2011