| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Division of Abdominal Imaging, University of Massachusetts Memorial Medical
Center, 55 Lake Avenue North, Worcester, MA 01655.
2 Division of Abdominal Imaging and Interventional Radiology, Massachusetts
General Hospital, Boston, MA 02114.
Received December 22, 2004;
accepted after revision November 17, 2005.
Address correspondence to A. K. Singh.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Fifty-seven patients (age range, 23-88 years) with abdominal collections who underwent imaging-guided catheter drainage were included in this retrospective study. These collections showed communication with either the pancreatic duct (n = 15) or the biliary duct (n = 42) on imaging. The imaging guidance included CT (n = 40), sonography (n = 17), and fluoroscopy (n = 4), either alone or in combination. The success of catheter drainage was described as resolution of the collection on follow-up imaging and clinical improvement. Other treatments directed toward management of leaks or collections were also recorded.
RESULTS. The success rates of catheter drainage for abdominal collections with biliary and pancreatic ductal communication were 93% (39/42) and 67% (10/15), respectively. The difference was statistically significant (p = 0.01). The three complications seen in this study were pneumothorax in one patient, bowel perforation in one, and death in one. The drainage catheter was upsized in five patients and an additional catheter was needed in nine patients.
CONCLUSION. Imaging-guided catheter drainage is a clinically useful option for drainage of abdominal collections that have pancreaticobiliary communications. The success rate is significantly better for collections with biliary communication than for those with pancreatic communication.
Keywords: biliary system • catheters • drainage • gallbladder • interventional radiology
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although a significant proportion of postpancreatitis collections resolve spontaneously, they may persist to form a pancreatic pseudocyst or they may be complicated by infection, necessitating drainage. Although surgery has been the treatment of choice in the past, it is now a second-line treatment [1]. Similarly, bilomas are also usually managed without surgery, either by an endoscopic biliary stent, nasobiliary drainage, percutaneous biloma aspiration, or catheter drainage [4].
Extravasated bile and exocrine pancreatic secretions irritate and inflame mesenchymal tissues. Consequently, extraluminal collections communicating with the biliary tree or the pancreatic duct offer unique challenges for percutaneous catheter drainage. In this study we determined the efficacy of percutaneous imaging-guided catheter drainage with or without adjunctive endoscopic interventions for abdominal collections that had a documented communication with either the biliary ducts or the pancreatic duct on ERCP, cholescintigraphy, or contrast injection studies.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
|
|
|
|
|
Additional catheters were used when the collections were either multiloculated or multiseptated or the initial catheter did not entirely drain the primary collection. Catheter upsizing was performed when the initial catheter was well positioned in the collection and still had low drainage rates of less than 10 mL per day.
Fifty patients had follow-up CT after the percutaneous procedure. One patient had follow-up ERCP that revealed healing of the pancreatic duct communication. Four patients underwent follow-up catheter injection studies, and one had a follow-up sonography evaluation. One patient had post-hepatectomy biloma communicating with the right hepatic duct and was clinically followed until the symptoms resolved.
Sixty-seven catheters were initially placed in 57 patients. All 67 catheters placements were technically successful. These were 7-French (n = 1), 8.5-French (n = 15), 10-French (n = 26), 12-French (n = 18), and 14-French (n = 7) catheters.
Data Collection and Analysis
Success of catheter drainage was described as resolution of the collection
on follow-up imaging and clinical improvement. Resorting to surgery after an
adequate trial of percutaneous catheter drainage was considered a failure of
percutaneous treatment. Surgery was performed if the symptomatic fluid
collection did not improve with the catheter trial over at least 1 week, or
when the patient's clinical condition worsened despite optimal percutaneous
management.
The fluid collections were evaluated for collection size, location, cause, culture results, and site of communication with the pancreatic or biliary ducts. Drainage success, duration of hospital stay, and the time between catheter placement and discharge from the hospital were compared. Complications related to imaging-guided drainage of the biliary and pancreatic collections were compared.
The chi-square test was used to compare the differences in the success rate, duration of hospital stay, culture positivity, and duration between catheter placement and discharge from the hospital. A p value of less than 0.05 was considered significant.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Causes
The cause of pancreatic collection in the five patients in whom catheter
drainage was not successful included a postsurgical collection in three, a
postpancreatitis collection in one, and a posttraumatic collection in one. The
three patients in whom biloma catheter drainage failed to resolve the
collection had postsurgical collections after partial hepatectomy (n
= 2) and after Whipple surgery (n = 1).
The most common cause of the fluid collection in the pancreatic drainage group was acute pancreatitis, which was seen in eight patients. The cause of the fluid collection in 15 patients was complicated pancreatitis in six, pancreatitis followed by trauma in two, and surgery (two Whipple surgery, two distal pancreatectomy, two necrosectomy, and one splenectomy) in seven patients.
In the biliary drainage group, the most common pathology of the fluid collection was recent cholecystectomy in 16, blunt trauma in six, pancreatitis in one, liver transplant in three, hepatic resection in seven, abdominal aortic aneurysm repair in one, hepaticojejunostomy in one, infected ventriculoperitoneal shunt in one, Klatskin tumor in one, Whipple surgery in three, and unknown cause in two patients.
Culture Results
The aspirated fluid cultures were positive in 52% (22/42) of patients in
the biliary drainage group and 60% (9/15) of patients in the pancreatic
drainage group. The difference in the frequency of positive cultures was not
statistically significant (p = 0.6). Eighteen (14 biliary and four
pancreatic) of the 57 patients in this study showed multiple bacteria on the
fluid culture.
In the biliary drainage group, the most common bacteria cultured included Enterococcus species in nine patients, coagulase-negative Staphylococcus species in four, Enterobacter organisms in three, and Escherichia organisms in five patients. In the pancreatic drainage group, the most common bacteria cultured included Enterococcus species in four patients, Bacteroides in one, Escherichia in one, Klebsiella in one, and Staphylococcus in one.
The patients who underwent biliary drainage had a catheter drainage success rate of 95% (21/22) in those with culture-positive collections and 90% (18/20) in those with culture-negative collections (p = 0.5). The success rates in the pancreatic drainage group were both 67% (6/9 and 4/6, p = 1) for culture-positive and culture-negative collections.
Pancreatic and Biliary Communication Sites
The site of communication of the biliary collection with the bile duct was
in the cystic duct in 10 patients (23.8%), the common bile duct in 10 (23.8%),
the right-sided intrahepatic duct in four (9.5%), the left-sided intrahepatic
duct in four (9.5%), and the right- and left-sided intrahepatic ducts in three
patients (7.1%). In the remaining 11 patients, the site of communication of
the collection was the Luscka's duct, a choledochoenteric anastomosis site,
the common hepatic duct, or the gallbladder. In the pancreatic collection
group, the site of communication was the main duct in 14 patients and a side
branch in one.
The three failures in the biliary drainage group and the five failures in the pancreatic drainage group occurred at the site of communication with the common bile duct and pancreatic duct, respectively. The site of communication of the collection with the pancreatic duct did not significantly correlate with successful outcome of the procedure (p = 0.5). The presence of common bile duct communication was a predictor of poor outcome of catheter drainage (p = 0.001).
Additional Interventions
Additional catheters for complete drainage of the collection were required
in nine (16%) of the 57 patients. In addition, five (9%) of the 57 patients
required catheter upsizing.
Six additional catheters were needed to drain the collection in three patients in the pancreatic drainage group, all of which were successfully drained. In two patients, the catheters were upsized because of incomplete drainage of the collection. The collections in all five of these patients were successfully managed by catheter drainage. The difference in the success between patients in whom additional catheter drainage or catheter upsizing was performed and patients in whom no additional interventions were performed was not statistically significant (p = 0.4).
In the biliary drainage group, 16 additional catheters were needed in six patients, all of whose collections were successfully managed by catheter drainage. Catheter upsizing was needed in three patients in this group and resulted in complete drainage of the collections. In one of the three patients in the biliary drainage group in whom catheter drainage failed, multiple additional catheters were unsuccessfully used.
Complications
The two major complications seen in this study were both in the biliary
drainage group; one was a pneumothorax that required catheter drainage, and
the other was a delayed duodenal perforation by a catheter in a patient with a
biliary-enteric anastomosis leak. In the latter, the bowel perforation was
caused by one of the two catheters and was successfully managed without
surgery. Two deaths occurred within 30 days among 57 patients in this study,
one each in the biliary drainage and pancreatic drainage groups. The single
death in the pancreatic drainage group occurred in a patient with pancreatitis
in whom both surgical necrosectomy and catheter drainage had failed. The death
in the biliary drainage group was caused by asystolic cardiac arrest in a
patient who had undergone recent Whipple surgery and was unrelated to the
biloma or the drainage procedure.
Duration of Hospital Stay
The durations of hospital stay and of percutaneous catheter drainage were
1-88 and 1-53 days, respectively. Table
1 indicates the duration of hospital stay and percutaneous
catheter drainage in the biliary and pancreatic drainage groups. The average
hospital stay in the biliary drainage group was 20.2 days and in the
pancreatic drainage group was 25.2 days. The time between imaging-guided
catheter drainage and discharge from the hospital was 15.1 days in the biliary
drainage group and 19.5 days in the pancreatic drainage group. These
differences were not statistically significant (p = 0.3 and 0.4).
|
Concurrent and Preceding Interventions
In the biliary drainage group, 60% (25/42) of patients had a preceding or
concurrent endoscopic intervention. These included a biliary stent and
nasobiliary drainage alone in 21 and three patients, respectively. In
addition, one patient had both a biliary stent and nasobiliary drainage, and
six of the 25 patients had percutaneous transhepatic biliary drainage. In the
pancreatic drainage group, a stent was used in five patients and
sphincterotomy was performed in one.
Percutaneous catheter drainage was successful in 24 of the 25 patients in the biliary drainage group and in five of the six patients in the pancreatic drainage group who had an additional endoscopic intervention. The success rates in patients with or without additional interventions were not statistically different in the biliary (p = 0.4) or the pancreatic (p = 0.4) drainage group.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Over the past two decades, the treatment of abdominal collections has changed from surgical to nonsurgical management using imaging-guided catheter drainage, stent placement, nasobiliary drainage, and antibiotic therapy. Imaging-guided catheter drainage of abdominal collections is associated with high cure rates [3, 4, 6].
Biliary ascites or localized bilomas are most often seen after biliary surgery and need nonsurgical management using imaging-guided catheter drainage or endoscopic procedures [7-9]. When managed surgically, bile duct injury with immediate repair, early reoperation, and delayed repair can be associated with serious complications in up to 31.8% patients [10]. In the study by Chen et al. [11], the success rate of nonsurgical modes of treatment of a postoperative bile leak was 82.5%. The mode of treatment in that study of 57 patients included percutaneous drainage, transhepatic biliary drainage, paracentesis, endoscopic management, and so forth [11]. In our study, we have shown the efficacy of percutaneous catheter drainage in treating patients with biloma who have a documented biliary collection. The procedure was successful in all 16 patients (100%) with a postcholecystectomy biloma and in 23 (88.5%) of the 26 patients with biloma from other causes. The most common site of postcholecystectomy leakage of bile was from the cystic duct stump; these leaks accounted for 50% of the cases.
Abdominal collections from a pancreatic cause are most often related to acute or chronic pancreatitis, as was also seen in this study. Acute pancreatitis and chronic pancreatitis result in pseudocyst formation in as many as 5% and 40% of patients, respectively [12]. If the pseudocysts are not infected or symptomatic because of their size, they can be closely observed. If pancreatic ductal communication is present, a pseudocyst can be managed initially with a transpapillary approach. Six of 15 patients in this group underwent an endoscopic intervention before percutaneous catheter drainage [13, 14]. Although in our study, the success rate of pancreatic drainage was lower than that of biliary drainage, the high rate of infection in pancreatic collections justifies the use of catheter drainage in our study.
In a study by Freeny et al. [15], the durations of catheter drainage for collections without and with pancreatic duct fistulas were 29 days and 96 days, respectively [15]. In general, the duration of catheter drainage in our study was shorter than those usually reported in the literature. This can be partly explained by the frequent use of a second catheter and by repositioning and upsizing of the catheters to maximize drainage. The aggressive management of catheters is part of our protocol for patient management, in which every patient is seen daily in the morning by an interventional radiologist, and problems of catheter nondrainage are urgently addressed.
Although the concurrent or preceding use of ERCP interventions, as seen in this study, might have an added or synergistic effect on the success of pancreatic or biliary collections, our study did not show any statistically significant difference between patients who received percutaneous drainage alone and those who received both percutaneous catheter drainage and adjunctive endoscopic intervention. The high rate of adjunctive endoscopic procedures in this study is because the diagnosis of pancreaticobiliary fistula is most confidently made on ERCP and the endoscopist often has the first opportunity to place a stent or nasobiliary drainage catheter. By decreasing the pressure head or removing the occlusion in a duct, endoscopic interventions may be helpful in decreasing fluid leaks from the biliary or pancreatic ducts and may promote healing, but successful drainage was achieved in some patients without an adjunctive endoscopic procedure. The retrospective nature of our study does not allow further conclusions regarding which patients are most likely to benefit from a combined endoscopic and percutaneous approach as opposed to percutaneous management alone.
We found a high incidence (52%) of infection in the biliary and pancreatic fluid collections. Bile, which is used in a number of culture media, inhibits the growth of many gram-positive and anaerobic microorganisms and promotes the growth of intestinal bacteria. Therefore, it is logical to see enteric flora as the predominant bacteria on culture.
The presence of infection in abdominal collections is a definite indication for a drainage procedure, because biliary stents, nasobiliary drainage, and sphincterotomy cannot arrest the disease process and the abscess must be cleared for a cure. The successful result of percutaneous catheter drainage of infected pancreatic fluid collections has ranged from 65% to 96% in the literature [15, 16]. Our results of pancreatic fluid drainage are similar to those reported by Freeny et al. [15], who showed a 65% rate of success for infected pancreatic fluid collections.
Imaging-guided catheter drainage is a clinically useful treatment option for the drainage of abdominal collections with pancreaticobiliary communication. It plays a major role in the management of postcholecystectomy bilomas. The success rate of imaging-guided catheter drainage is significantly better for collections with biliary communication than for those with pancreatic communication. However, for either condition, the success rate is sufficiently high to justify the use of percutaneous catheter drainage as the technique of choice in collections with pancreatic and biliary communication.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
