| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vascular and Interventional Radiology |
1 From the Department of Angiography and Interventional Radiology, University of Vienna Medical School, Waehringerguertel 18-20, A-1090 Vienna, Austria (M.S., A.S., M.A.F., M.C., J.L.); Department of Radiology, University of Rome "La Sapienza," Italy (P.R., M.B.); and Interventional Radiology, Medical University of South Carolina, Charleston (R.U.). Received October 25, 2001; revision requested December 3; revision received February 15, 2002; accepted March 14. Address correspondence to M.S. (e-mail: maria.schoder@univie.ac.at).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: This prospective nonrandomized study included 42 patients with malignant obstruction of the common bile duct, common hepatic duct, and hilar confluence. Unilateral (n = 38) or bilateral (n = 4) bile duct drainage was performed by using fully covered endoprostheses with anchoring fins. To avoid branch duct blockage, endoprostheses with drainage holes at the proximal end were available. Procedure- and device-related complications were recorded. Patient survival and stent patency rates were calculated with Kaplan-Meier survival analysis. Mean follow-up bilirubin and alkaline phosphatase levels were calculated, and differences in means were evaluated with a paired t test.
RESULTS: Successful deployment, correct positioning, and patency of the device were achieved in all patients. Procedure-related complications occurred in two (5%) patients. Thirty-day mortality rate was 20% (eight of 41 patients), and median survival time was 146 days. Laboratory values decreased significantly after the procedure (P < .001). Recurrent obstructive jaundice occurred in six (15%) patients. Primary patency rates at 3, 6, and 12 months were 90%, 76%, and 76%, respectively. Calculation of the composite end point of death or obstruction revealed a median patency duration of 138 days. No endoprosthesis migration was observed. Branch duct obstruction was observed in four (10%) patients. Postmortem examination of one stent revealed a widely patent endoprosthesis with intact covering.
CONCLUSION: Initial results of percutaneous treatment of malignant biliary obstructions with fully covered ePTFE-FEP endoprostheses suggest that they are safe and potentially clinically effective.
© RSNA, 2002
Index terms: Bile duct radiography, 76.1225, 76.1226 • Bile ducts, interventional procedures, 76.1267, 76.1269 • Bile ducts, neoplasms, 76.32 • Bile ducts, stenosis or obstruction, 76.32 • Bile ducts, stents and prostheses, 76.1267, 76.1269
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
To improve the patency rates achieved with metallic stents, a few studies (6,15–19) to investigate polyurethane-covered metallic stents have been performed. However, in these studies, the patency duration was not prolonged compared with the duration achieved with noncovered stents. In the three studies (6,15,19) in which polyurethane-covered metallic stents were used, each of the authors reported defects in the polyurethane covering, with tumor ingrowth and consequent stent obstruction; these findings suggest that polyurethane degrades over time. In animal studies conducted by the manufacturer of the endoprosthesis used in this study (W. L. Gore & Associates, unpublished data, 2000), the expanded polytetrafluoroethylene–fluorinated ethylene propylene (ePTFE-FEP)–covered stent was not degraded by bile at 3 months. Therefore, we hypothesized that covering a stent with ePTFE-FEP might improve the patency rates.
The purpose of our study was to determine and present the initial technical and clinical results of using an ePTFE-FEP–covered biliary endoprosthesis (W. L. Gore & Associates, Flagstaff, Ariz) for the treatment of malignant biliary obstruction.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
The primary indication for inclusion in the study was malignant obstruction of the common bile or hepatic duct, including the hepatic duct confluence, by a nonresectable tumor. Only patients who were aged 21 years or older were included, and written informed consent was obtained from all patients before they underwent the procedure. Exclusion criteria were history of previous biliary surgery, multiple strictures that required treatment, presence of a nonremovable metallic biliary stent, diagnosis of active infection of the biliary system, chronic liver disease, uncontrolled coagulation, severe allergy to contrast material, and/or poor clinical condition with an estimated life expectancy of less than 3 months.
Patients
The 42 patients in the study were 22 women and 20 men (mean age, 72.5 years; age range, 43–92 years). The diagnoses of tumor nonresectability were based on the results of explorative laparotomy or the findings of either ultrasonography (US), computed tomography (CT), magnetic resonance (MR) imaging, or endoscopic retrograde cholangiopancreatography in combination with those of histologic analysis of needle biopsy or brush biopsy specimens. All images were read by experienced abdominal radiologists, who in consensus with the involved abdominal surgeons established the diagnosis of nonresectability. Twenty-six patients had pancreatic carcinoma; three, cholangiocellular carcinoma; one, gallbladder carcinoma; and one, obstruction caused by cancer of the duodenum. In the remaining 11 patients, bile duct obstruction was caused by enlarged lymph nodes due to metastasis in 10 patients and by lymphoma in one patient.
According to TNM tumor staging criteria (20), 35 patients had a stage IV; four patients, a stage III; and one patient, a stage II tumor. For two patients, tumor classification data were missing. The proximal level of obstruction was graded according to the Bismuth system of classifying bile duct injury in relation to the confluence of the hepatic ducts (21). Thirty-two patients had a type 1 lesion—that is, the proximal end of the obstruction was more than 2 cm distal from the hepatic bifurcation; six, a type 2 lesion—that is, the proximal end of the obstruction was within 2 cm distal from the bifurcation; and four, a type 4 lesion—that is, the obstruction completely obliterated the bifurcation. At admission, nearly all patients underwent complete laboratory assessment (Table 1).
|
|
Endoprosthesis Placement Procedure
All procedures were performed by one of four authors (M.S., P.R., R.U., M.B.). Initially, to evaluate the bile duct anatomy, a standard percutaneous transhepatic cholangiogram was obtained in all patients, and based on the findings, the final determination of eligibility for study participation was decided (Fig 2). Puncture of the intrahepatic biliary tract was performed unilaterally in 38 patients and bilaterally in four patients (one patient with Bismuth type 2 lesions, and three patients with Bismuth type 4 lesions). The obstructing lesions were traversed by using standard guide wire techniques. Conventional external-to-internal drainage was achieved with a multi–side-hole catheter during the first session. Flushing the catheter with contrast material enabled accurate delineation of the lower level of the obstructing lesion. In a second session 2–5 days after the first procedure, the drainage catheter was removed over a 0.035-inch extra stiff guide wire (Amplatz; William Cook Europe, Bjaeverskov, Denmark) or a stiff hydrophilic wire (Terumo, Tokyo, Japan), and a 9- or 10-F introducer sheath was inserted. Before insertion of the endoprosthesis, dilation of the stricture was performed with a low-profile, 5–10-mm high-pressure angioplasty balloon in 20 patients.
|
|
|
The procedure was performed with antibiotic prophylaxes, which were administered intravenously in 38 patients and orally in four. All patients continued antibiotic treatment for 5–6 days after the initial procedure.
Follow-up and Statistical Analysis
Data regarding the technical aspects of stent implantation—namely, delivery system trackability with the guide wire, deployment accuracy, radiographic visibility, and device withdrawal through the sheath—and the imaging findings after deployment—namely, correct positioning and patency of the device and diameter of the endoprosthesis immediately after deployment—were recorded. Furthermore, information regarding predilation and postdilation of the strictures and the procedure- and device-related complications were recorded as part of the study protocol. Imaging studies (ie, US, CT, and cholangiography) were performed in patients who developed pain or recurrent symptoms of jaundice.
The means and standard errors of the means for all laboratory values were calculated, and the significance of differences in the means was evaluated by using a paired t test. A P value of less than .05 indicated a significant difference.
The follow-up status of each patient was based on laboratory-tested levels of bilirubin and alkaline phosphatase and on clinical symptoms evaluated 1, 3, and 6 months after the stent placement procedure. After 6 months, all patients were then followed up until recurrence of obstructive jaundice or death. If a patient could not be seen at the clinic, follow-up data were requested by means of communication with the general practitioners or referring physicians. The observation period for this study was terminated at the end of the 6-month follow-up of the last patient in whom a stent was implanted. Endoprosthesis patency and patient survival were calculated by using Kaplan-Meier survival (life-table) analysis. The duration of primary patency of the endoprosthesis was defined as the interval between stent placement and recurrence of obstructive jaundice. If obstruction was not evident during a patient’s life, the patency period was considered to be equal to the survival duration but censored. The results of any postmortem examinations performed during the study period were sought.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Initially, an 8-cm-long endoprosthesis was implanted in 32 patients, a 6-cm-long endoprosthesis was implanted in eight, and a 4-cm-long endoprosthesis was implanted in two. To relieve obstruction approximately 2 cm at the proximal and distal ends, a total of 57 endoprostheses were necessary. Two endoprostheses were inserted in eight patients: five patients with Bismuth type 1 lesions, two with Bismuth type 2 lesions, and one with missing lesion classification data. One patient with a type 4 lesion needed three stents for adequate drainage, and one patient with bilateral drainage needed six stents for adequate drainage of the length of the obstruction. The proximal end of the endoprosthesis was located in the common bile duct in 10 patients, in the common hepatic duct in 27 patients, in the right hepatic duct in one patient, and in both hepatic ducts in four patients. On the distal side, the endoprosthesis was located in the duodenum in 22 patients, at the level of the papilla in seven patients, and in the common bile duct in 13 patients. Twenty-six of the inserted endoprostheses were 8 mm in diameter, and 31 were 10 mm in diameter.
After deployment of the endoprosthesis, the mean minimum diameter along the stricture was 7.2 mm (range, 3–10 mm). In 19 patients, stent expansion was inadequate—that is, less than 50% of the original diameter—after deployment, so dilation with balloon catheters with diameters of 6–10 mm was performed. In all but four patients, the nominal maximum diameter of the endoprosthesis was reached proximal and distal to the stricture immediately after deployment.
Complications
Two (5%) patients had procedure-related complications. One patient developed a perihepatic bile leak, which was successfully treated with percutaneous drainage. In the other patient, perihepatic and intrahepatic hematomas that were diagnosed 5 days after the procedure required percutaneous drainage but not blood transfusion.
Four (10%) patients had device-related complications that were caused by obstruction of duct branches due to the endoprosthesis. One patient had mild symptoms of pancreatitis that began 3 days after the procedure and resolved completely with conservative therapy within 10 days. Three patients had symptoms of acute cholecystitis; two of these patients were successfully treated with cholecystostomy, and one underwent surgical cholecystectomy.
Patient Survival and Stent Patency
One patient, who was homeless, was lost to follow-up and consequently was not included in the Kaplan-Meier survival analysis. Eight (20%) patients died within 30 days after the percutaneous treatment. Of these patients, one had perforation of the small bowel caused by erosion of tumor, one had a massive pulmonary embolism, and one had septic multiorgan failure due to a left subphrenic abscess after explorative laparotomy performed before percutaneous treatment. In five patients, death was caused by advanced cancer and poor clinical condition and was also not directly related to the drainage procedure.
During our period of review, 26 (63%) of 41 patients died with a primary patent stent, and nine patients with primary patent stents were alive (range, 172–395 days) (Table 2, Fig 3) at the time this article was written. According to the Kaplan-Meier survival analysis results, the median survival time was 146 days (standard error, 33) (mean survival, 173 days ± 22).
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
A potential strategy to prevent tumor ingrowth is to cover stents with plastic materials, such as silicone, polyether-polyurethane, and polycaprolactone (24). In an experimental study, stainless steel stents covered with one of these three polymeric compounds were placed in the common bile duct in 18 dogs. Yasumori et al (24) found that polyurethane and silicone had a low coefficient of friction and therefore were durable polymers, whereas polycaprolactone was fragile and thus not considered a suitable covering. Microscopically, these investigators found minimal inflammatory cell infiltration in the perimucosal connective tissue. Nevertheless, luminal narrowing caused by mucosal hyperplasia at both ends of the stents had occurred in all animals after 3, 6, and 12 months.
Similar hyperplastic changes have been noted in canine biliary tracts following placement of noncovered and covered self-expandable stents after 1 month (25). On the other hand, this study group found that mucosal hyperplasia, as well as submucosal fibrosis, was more marked with noncovered stents than with silicone-covered stents. The authors concluded that the covering prevented embedding of the wires in the mucosa, and, therefore, the degrees of hyperplasia and fibrosis were decreased. Mucosal hyperplasia is uncommon in humans, however. Boguth et al (12) found no mucosal hyperplasia in 22 autopsy specimens after implantation of noncovered self-expanding stents; the survival period in that study ranged from 5 to 575 days. In our autopsy specimens, there was no sign of hyperplasia. In addition, the inner luminal surface was free of bile sludge and the interstices of the implant showed no evidence of neoplastic cellular infiltration. Because of these histologic findings 30 days after implantation of the endoprostheses, we speculate that bacterial adherence, as well as tumor ingrowth, could be prevented or decreased with use of the ePTFE-FEP covering.
The obstruction of stents with biliary sludge is another problem. Preliminary data regarding stent clogging have shown that bacterial colonization on the stent surface is the basis for sludge formation (26–28). Normally, the biliary tracts in humans (29) and rodents (30) are sterile. In studies with implanted biomaterial, indwelling devices were recognized as foreign bodies by the host and were coated with a film of proteins (31). Furthermore, this biofilm enables bacteria to persist on the surface of the implant and protects bacteria from bactericidal agents.
However, the results of in vitro studies performed by Tsang et al (32,33) have shown that the perfusion of polyethylene- and silicone-covered stents with ampicillin-sulbactam prevented bacterial adherence and biofilm formation and therefore prolonged stent patency. To our knowledge, to date, these laboratory findings have not been confirmed by clinical studies. Results of a prospective randomized study performed by Sung et al (34) failed to show prolonged stent patency in patients treated with polyethylene stents and long-term prophylactic protection with ciprofloxacin versus the patency in those treated with stent placement and no antibiotics. Therefore, efforts have been undertaken to evaluate materials that are effective in reducing bacterial colonization (35,36).
In vitro studies performed by the manufacturer of the endoprosthesis that we used in the current study demonstrated that nonporous ePTFE-FEP is more resistant to bacterial attachment than is standard expanded polytetrafluoroethylene. Furthermore, animal study results have shown that the expanded polytetrafluoroethylene covering of the stents is not affected by bile at 3-month follow-up (W. L. Gore & Associates, unpublished data, 2000). Therefore, we expected diminished biofilm formation and the resulting prevention of stent clogging by bile sludge. In the present study, none of the endoprostheses became reobstructed by bile sludge. Nevertheless, further studies are required to strengthen these results in clinical practice and to evaluate whether ePTFE-FEP is superior to polyurethane as a cover material.
A short median survival time is characteristic of patients who present with malignant obstructive jaundice; most reports (1–4,6,12,15,23,37) describe an average survival time of less than 9 months. This makes it difficult to compare the patency rates with different stents. Most patients die before their stents have been in place long enough to have become occluded. There has been a considerable range of reported rates of recurrent obstruction after implantation of covered and noncovered metallic stents. Lammer et al (22) treated 53 patients who had malignant biliary obstructions with self-expanding stainless steel stents, and only six (11%) of these patients experienced recurrent obstructive jaundice. The mean observation period, however, was 4.5 months. Boguth et al (12) observed recurrent obstructive jaundice in 20% (n = 12) of 59 patients after treatment of malignant obstructions with placement of self-expanding stainless steel stents; the mean survival time in their series was less than 6 months. A short survival period precludes a meaningful assessment of stent patency.
O’Brien et al (14) treated 28 patients with malignant obstruction by means of endoscopic implantation of self-expanding stainless steel stents. During a survival time of 1.0–38.5 months (median, 15.1 months), they observed a 46% (13 patients) reocclusion rate. Lee et al (37) observed 25- and 50-week patency rates of 81% and 53%, respectively, in a study cohort of 100 patients treated with various metallic stents. In a European multicenter study that included 240 patients (13), significant differences in the patency rates of four types of metallic stents were observed. Self-expandable nitinol and Ni-Co-Ti alloy stents had 25-week patency rates of 78% and 67%, respectively, whereas Z-stents and balloon-expandable tantalum stents had 25-week patency rates of 30% and 20%, respectively.
There are publications on the treatment of malignant biliary obstructions with self-expandable metallic stents covered with polyurethane. According to the report by Miyayama et al (16), who used partially covered Z-stents, only one (5%) occlusion occurred in one of 19 patients owing to tumor overgrowth, which was proven at cholangiography. In the study performed by Kanasaki et al (18), which involved 18 patients who were treated with placement of a custom-fabricated polyurethane-covered nitinol stent, two (11%) patients had recurrent obstructive jaundice. A 30%–67% rate of recurrent obstructive jaundice was reported in three series (6,15,19) in which partially polyurethane-covered, self-expanding stainless steel stents were implanted. In addition to obstruction by bile sludge, neoplastic tissue was found inside the stent. These authors suggested that defects in the covering owing to mechanical factors during deployment could have been responsible for the tumor ingrowth. They also suspected that the covering may not have been as biostable as assumed and may have been dissolved in the duodenal and gastric juices over time.
In our study, the obstruction rate was 15% (six of 41 patients). The median survival time, however, was 146 days. We suspect that the extension of the endoprosthesis beyond the stenosis and into the normal duct above the obstruction was inadequate, and, therefore, tumor overgrowth caused the stent obstruction. However, our thesis was not proven histologically. To overcome this problem, a so-called overstenting of the obstruction by at least 3 cm was recommended by Kanasaki et al (18), who did not observe occlusions owing to tumor overgrowth.
In our study with ePTFE-FEP–covered stents, the patency rates with the endoprostheses at 3, 6, and 12 months were 90%, 76%, and 76%, respectively. The patency rate at 6-month follow-up was similar to the 79% patency rate reported with covered Z-stents (16), as well as to those observed with some types of noncovered self-expandable stents (13,37). In other series, in which polyurethane-covered self-expandable stainless steel stents were used to treat malignant biliary obstructions (6,15), patency rates of 47% at 6-month follow-up were reported. When the results reported by Hausegger et al (6) and Rossi et al (15) are compared with the patency rates in our study, it appears that ePTFE-FEP–covered stents tend to yield higher patency rates at midterm follow-up.
Occlusion of the cystic and pancreatic ducts by the covering are a potential risk (6,15,17). However, Kanasaki et al (18) found no symptoms of cholecystitis or pancreatitis in patients, despite cystic and pancreatic obstruction by the covering. They assumed that chronic obstruction by the tumor caused the absence of these complications. In some of our study patients, because of tumor localization, obstruction of the cystic and pancreatic ducts could not be avoided. One patient, who initially had a nonimpaired pancreatic duct, later developed mild symptoms of pancreatitis, and three patients had to be treated because of symptoms of cholecystitis. The risk of cystic duct blockage could be minimized by implanting an endoprosthesis with perforations at the proximal end of the covering, similar to the biliary endoprosthesis used in our study. Moreover, this design feature allows stent placement above the bifurcation without functional obstruction of the segmentary intrahepatic ducts. In addition to protecting duct branches, side holes increase the potential for bacterial attachment and biofilm formation (27).
Fully or partially covered stents seem to be prone to migration, which has been observed in 3%–11% of cases (6,15–17,19). However, the endoprostheses used in our study have anchoring fins at each end, and none of the stents migrated during the observation period. The basic limitation of our study is that it was designed as a feasibility trial; thus, further randomized studies will be necessary to compare ePTFE-FEP–covered endoprostheses with bare stents or other stent-grafts.
In conclusion, the placement of ePTFE-FEP–covered biliary endoprostheses for treatment of malignant biliary obstructions can be considered safe and effective, and the anchoring mechanisms prevent stent migration. To obtain sufficient knowledge about the biostability of the membrane, which could influence the rate of tumor ingrowth–related obstructions, further evaluations of the endoprosthesis are necessary.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, M.S.; study concepts and design, all authors; literature research, M.S.; clinical studies, all authors; data acquisition and analysis/interpretation, all authors; statistical analysis, M.A.F., A.S.; manuscript preparation, definition of intellectual content, and editing, M.S.; manuscript revision/review, J.L.; manuscript final version approval, M.S., J.L.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Avritscher, D. C. Madoff, P. T. Ramirez, M. J. Wallace, K. Ahrar, F. A. Morello Jr, S. Gupta, R. Murthy, K. C. Wright, and M. E. Hicks Fistulas of the Lower Urinary Tract: Percutaneous Approaches for the Management of a Difficult Clinical Entity RadioGraphics, October 1, 2004; 24(suppl_1): S217 - S236. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
