Rungsun Rerknimitr

Microbiology of bile in patients with cholangitis or cholestasis with and without plastic biliary endoprosthesis

BACKGROUND Cholangitis is a frequent complication of biliary stents. Empiric antibiotic selection is primarily based on bile cultures obtained from patients undergoing surgery and few data are available with respect to ERCP and biliary stents. This study analyzed the microbiology of bile and the antibiotic sensitivities of the organisms identified in patients presenting with clinical cholangitis or cholestasis. METHODS All bile cultures collected during ERCP from January 1994 to January 2000 were identified by using an ERCP database. RESULTS One hundred eighty cultures from 160 patients (mean age 55 years, range 6-94 years) were identified. Sixty-nine specimens were collected at ERCPs performed in patients with no biliary stent in place (group 1). One hundred eleven specimens were obtained from patients with a biliary stent in situ (group 2). The primary diagnoses for patients in groups 1 and 2 were as follows, respectively: malignant bile duct obstruction (25 and 85), choledocholithiasis (16 and 2), benign biliary strictures (6 and 18), acquired immunodeficiency syndrome (8 and 0), and miscellaneous (14 and 6). There were 38 positive bile cultures in group 1 (55%) and 109 positive cultures in group 2 (98%; p < 0.05). A higher frequency of polymicrobial infections (90%) was found in group 2 patients (vs. group 1, 45%; p < 0.001). Escherichia coli was the most common organism found in group 1 (17%); Enterococcus the most common in group 2 (31%). Among patients with positive bile cultures, bacteremia was more frequent in group 2 patients (46% vs. 21%; p < 0.05). Ciprofloxacin and cefiriaxone were the most effective antibiotics against identified gram-negative bacilli, and vancomycin against Enterococci. CONCLUSION In patients with cholangitis associated with biliary obstruction, the antibiotic selected initially should be active against gram-negative bacilli. Use of quinolones is recommended because these agents effectively penetrate an obstructed biliary tree and can be administered orally. Enterococci and polymicrobial infections are found more commonly in patients with a biliary stent than those without a stent. Pending definitive biliary decompression, patients with sepsis and those who do not quickly respond to treatment with a quinolone may benefit from the addition of antibiotic coverage against gram-positive organisms, targeted against Enterococci.

3802 Pancreatic duct leaks: results of endoscopic management.

Pancreatic duct leak (PL) is a complication of acute or chronic pancreatitis or pancreatic trauma. Surgical therapy(Rx) has been the traditional Rx for PLs that fail to resolve with medical Rx. Recent reports support a role for endoscopic therapy (ERx). We report the result of ERx in pts with PL. Methods: From 4/94 to 10/99, 153 pts (91 men, 62 women) mean age 51 yrs (range 9-89) were found to have PL at ERP. According to ERP finding, pts were categorized as 1) complete pancreatic duct disruptions (CPDD)-Main PD disruption without visualization or wire passage into upstream duct 2) partial main pancreatic duct disruption (PPDD)- Main PD visualized upstream from PL and 3)small duct leak (SD)-Leak from side branch. ERx included endoscopic sphincterotomy (ES), pancreatic duct stricture dilation if needed, and pancreatic stent or nasopancreatic catheter (NPC) placement. Pts with an associated accessible pseudocyst had endoscopic drainage by a transpapillary, transmural, or combined approach. Longterm follow-up was obtained by CT, ERP, clinical evaluation, and phone calls. Results: See table. Summary: 1) ERx of PL was successful in 102/121 (84%) pts with a recurrence rate of 4%. 2) pts with PPDD (98%) or SD (100%) were more likely to seal than pts with complete disruptions (62%, p

Does stricture dilation add to the yield of brush cytology in the evaluation of malignant biliary obstruction (MBO)

Does stricture dilation add to the yield of brush cytology in the evaluation of malignant biliary obstruction (MBO)?

4603 Spontaneous dislodgment rates of long length, small diameter, “unflanged” pancreatic stents placed partially up the duct.

Pancreatic duct (PD) stent placement reduces pancreatitis rates following precut biliary sphincterotomy (ES) or pancreatobiliary ES in pts with sphincter of Oddi dysfunction (SOD). To limit ductal injury and need for endoscopic removal, spontaneous dislodgment is desirable after 1-2 weeks. Short (=3cm) unflanged stents dislodge 80% of the time. However, post- ERCP pancreatitis rates remain relatively high, ~10-15% in SOD pts. We hypothesized that these short stents might dislodge too quickly (ie

4611 Does placement of a small diameter, long length, unflanged pancreatic duct stent reduce the incidence of post-ercp pancreatitis?

Background: Temporary pancreatic duct (PD) stent placement reduces pancreatitis rates following needle-knife pre-cut biliary sphincterotomy (ES) or pancreatobiliary ES in pts with sphincter of Oddi dysfunction (SOD). Such stents have typically been 4-5F diameter and 2cm long.We hypothesized that a small diameter stent would cause less ductal irritation, but longer lengths might prevent premature dislodgment. Small diameter (3F), long length (8-10cm) unflanged PD stents have a high spontaneous dislodgment rate (83%), and they appear to be infrequently associated with PD injury. It has been suggested that the 3F luminal diameter may obstruct prematurely, resulting in increased pancreatitis rates. This study addresses this concern. Methods: From 1/94-10/99, 1087 pts had a 3/4-single pigtail unflanged PD stent (Wilson-Cook, Winston- Salem, NC) placed at ERCP for a variety of conditions. Stent diameter ranged from 3-7F. All 3F stents were 8-10cm long, while the other stents were 2-10cm, with a variable length remaining in the duodenal lumen. Results: See Table. Summary: 1.Pancreatitis rates remain relatively high (13.8%) in these high-risk patients. 2. Placement of a small diameter (3F), unflanged PD stent is no less effective than larger diameter stents in prevention of pancreatitis. Moderate to severe pancreatitis rates are lower than historical controls. Conclusion: Prospective randomized trials comparing 3F and larger diameter stents (of varying lengths) would be of interest. Further study addressing PD and parenchymal injury with these stents is ongoing.

3566 Endoscopic therapy of biliary and pancreatic disorders in children.

There is limited data on the role of ERCP in the pediatric population. We report our experience in this group during the last 5 years. Pts and Methods: From 1/94-9/99, 126 pediatric pts were evaluated by ERCP. There were 47 boys and 79 girls with a mean age of 12.1 yrs (range 0.1-17). There were 170 ERCPs performed (43 diagnostic and 127 therapeutic). Ninetytwo procedures were performed under general anesthesia and the remainder were under conscious sedation. A pediatric endoscope (diameter 8.5 mm) was used in 9 pts(all less than 2 yr)and a standard adult diagnostic endoscope (diameter 11 mm) was used in the remainder. Indications for ERCPs were: Idiopathic acute pancreatitis (n=15), recurrent acute pancreatitis (n=31), chronic pancreatitis (n=8), obstructive jaundice (n=40), suspected sphincter of Oddi dysfunction (SOD) (n=22), and miscellaneous (n=10). The desired duct was visualized in 116 of 126 pts (92.1%). Results: Therapeutic techniques from 170 sessions are shown on the table. Disease findings were (some pts had more than one finding): bile duct stone(s) (n=18), SOD (n=24), biliary atresia (n=1), anomalous P-B junction (n=3), choledochal cyst (n=4), biliary fistula (n=5), PSC (n=5), chronic pancreatitis with strictures and stones (n=11) or without stones (n=5), pancreas divisum (n=18), pancreatic duct leak (n=5), benign biliary stricture (n=7), malignant biliary stricture (n=1) and normal (n=12). Twenty-four complications occurred in 21 pts (16.6% of 126 pts). Twenty pts developed acute pancreatitis graded mild in 14, moderate in 3, and severe in 3. Conclusion: ERCP is an effective diagnostic tool in the evaluation of biliary and pancreatic disorders in the pediatric population. It offers therapeutic application that can obviate surgery with an acceptable complication rate similar to the adult population.

⁎4635 Post orthotopic liver transplantation choledocholithiasis: endoscopic findings and results.

In OLT patients, stone formation appears to occur at an accelerated rate compared to the general population because of associated recurrent biliary tract infection, epithelial shedding, cyclosporine use and biliary strictures. We report the endoscopic findings and treatment results in OLT patients with a duct-to-duct anastomosis (DDA). Patients and Methods: From 5/88- 8/99, 412 OLTs were performed and 121 OLT patients with DDA underwent 325 ERCPs for evaluation of cholestasis. All duct stone patients underwent attempted endoscopic stone extraction following endoscopic sphincterotomy (ES), biliary stricture dilation or stone lithotripsy if needed. Results: Forty-six patients were found to have stones (n=37)or sludge (n=9). Stones were identified in 18 patients on T-tube cholangiogram prior to ERCP. Twenty-eight patients presented with cholangitis. Bile duct stones/sludge were detected at a mean of 19.2 months (range 1.8-53.5 months) after transplant. Twenty-nine patients also had biliary strictures and 28 had stone(s) upstream to the stricture. Eleven patients had castlike stones. Stones were located in both the donor and recipient ducts in 14 patients, in the donor duct alone in 18, and in the recipient duct alone in 14. Stones were present in intrahepatic ducts in 7 patients. ES was performed in all cases for stone and sludge removal. Stone removal was successful in all 46 patients and required one ERCP session in 27 patients (58.7%), two sessions in 11 patients (23.9%), and three or more sessions in 8 patients (17.4%). All patients with stones and stricture required at least 2 sessions of endoscopic treatment which included stent placement for stricture management. Recurrent stones were found in 8 patients (17.4%) during follow-up ERCP at a mean of 5.6 months (range 1.9-18.7 months). Conclusion: Endoscopic stone removal in OLT patients with duct-to-duct anastomosis can be performed in the same manner as the general population with choledocholithiasis. However, multiple sessions may be required, especially in patients with an associated stricture. Medical therapy to prevent stone formation/recurrence is needed.

⁎⁎3801 Complete disruption of main pancreatic duct: short and long term results of endoscopic management.

Historically, surgical management has been the definitive treatment in pts with CPDD.We report short and long term results of ERx in this group of pts. Pts and Methods: From 4/94-10/99 153 pts (91 men, 62 women; mean age 51 yrs, range 9-89) were found to have a pancreatic duct leak at ERP. 37 pts had a CPDD. Indications for ERP were: pseudocyst (n=14), pancreatic ascites (n=6), unresolving pancreatitis (n=11), pancreaticocutaneous fistula (n=6). CT findings were: pseudocyst (n=14), ascites (n=6), disconnected tail (n= 3), pancreatic edema/necrosis (n=9). Five pts were sent directly to surgery. Thirty-two pts underwent 84 sessions of ERx which included endoscopic sphincterotomy, pancreatic duct stricture dilation if needed, and stent or nasopancreatic catheter placement. Pts with an associated accessible pseudocyst had endoscopic drainage by transpapillary stenting (n=6), cystogastrostomy (n=3), cystoduodenostomy (n=1), or combined procedures (n=1). Long-term follow-up was obtained by CT, ERP, clinical evaluation, and phone calls. Results: Twenty pts initially sealed their leaks at a mean duration of 1.6 month (range 0.2-4.5). Twelve pts required surgery after failed ERx. Pts with an initial sealed leak were evaluated at follow-up ERP with or without CT scan and categorized into 1) Successful upstream duct (UD) reconnection (SUR) and 2) Duct obstruction (DO) at prior leak site. Results of a 30-month follow-up are shown on table. Pts were asymptomatic (n=9), had recurrent pancreatitis (n=3), chronic pain (n=2), recurrent fistula (n=2), or were unable to be contacted (n=4). Summary: CPDD closure rate was 63% (20/32); 10% had a confirmed recurrent duct leak (fistula) 2) 25% continue to have pancreatic symptoms. Conclusion: ERx appears to be an acceptable alternative to surgery for a subset of pts with CPDD. Long-term results of pts with SUR are excellent. In pts with DO, the undrained segment may remain asymptomatic, become atrophic, or be associated with recurrent pancreatitis or recurrent fistulization. Better methods are needed to select the subset of pts with CPDD who will respond to ERx.

⁎4682 Post orthotopic liver transplantation bile leak: results of endoscopic management.

Bile leak is one of the most common biliary complications after OLT. Because of the relatively high morbidity and mortality rates associated with surgical intervention, nonsurgical methods of treatment are being investigated. Endoscopic drainage has become a popular approach in OLT pts with a duct-to-duct anastomosis (DDA).We report our results of endoscopic therapy (ERx) in these pts. Patients and Methods: From 5/88-8/99, 408 OLTs were performed at our institution and 4 pts who had OLT at another hospital were followed by us. Twenty-two were referred for evaluation of a suspected bile leak. ERx included endoscopic sphincterotomy (ES), stent or nasobiliary tube (NBT) placement. Resolution of the bile leak was confirmed either by repeat ERCP or NBT cholangiogram. Results: A bile leak was diagnosed in 22 pts (15 men, 7 women; mean age 37.8 years, range 9-51 years). The mean interval between the OLT and endoscopic intervention was 9.8 weeks (range, 1-20 weeks). This complication occurred in 6.0% of the total number of pts who received transplants and had a DDA during this period. ERCP demonstrated a leak in all 22 pts arising from the T-tube tract (n=13), anastomosis (n=7), intrahepatic radicle (n=1), donor cystic duct (n=1). Six pts had fever at presentation and another six showed evidence of leak per T-tube cholangiogram. Management of the bile leak is shown on the table. Pts undergoing stent therapy had repeat ERCP with stent removal (if the leak sealed) at a mean time of 32 days (range, 22-60 days). Two pts who underwent NBT had fistula closure on day 6 and 8. Two pts also had stones removed at the time of stenting. All leaks treated by endoscopy sealed clinically and confirmed by follow-up ERCP or cholangiogram via NBT. Endoscopic evaluation and Rx was associated with a 4.5% major complication rate and 0% mortality rate. Summary: ERCP demonstrated a biliary fistula in 22 of 22 pts (100%). Endosocpic management resolved biliary fistulae in all pts in whom it was attempted. Conclusion: Endoscopic management of bile leak in OLT pts with DDA is very effective and can obviate the need for surgical intervention. Choice of biliary decompression does not effect the outcome of this approach.

7239 Stricture dilation and brush cytology in the evaluation of malignant biliary obstruction (mbo): two brushes are better than one.

Background: Tissue confirmation is desirable in pts with suspected MBO. Endoscopic manipulation of the stricture prior to tissue sampling may increase tumor exfoliation, making more malignant cells available for diagnosis. This on-going study compares the yield of brush cytology before and after stricture dilation applied at ERCP. Methods: Pts found to have a biliary stricture at ERCP suspicious for neoplasia initially underwent brush cytology with a standard Geenen brush (Wilson-Cook). Following stricture dilation (pneumatic balloon and/or Soehendra dilating catheter), repeat brush cytology was obtained. Final cancer diagnosis was based on cytologic results plus surgery, EUS, autopsy or clinical follow-up. Results: From 6/98-11/99, 59 pts had 60 double brush cytology specimens obtained from 41 malignancies (42 specimens): 26 pancreatic, 9 cholangioCA, 3 metastases, 2 gallbladder, 1 lymphoma. If highly atypical cells were considered as cancer, 15/42 pts (35.7%) had cytologic brushings positive for malignancy. However, only 6 pts were found to have malignant cells using both brushes. 3 pts (2 panc CA, 1 cholangioCA) had positive cytology with the pre-dilation specimen but only benign cells post-dilation. 6 pts (4 panc CA, 2 cholangioCA) had negative cytology pre-dilation but malignant cells post-dilation. Subgroup analysis revealed no difference between the balloon and Soehendra dilating catheters. Summary: There were no differences in sensitivity for cancer detection pre- and post-stricture dilation (21.4% vs. 28.6%, p=NS). The addition of a 2nd brush cytology specimen did increase sensitivity to 35.7% (p=.024). Conclusions: The cancer detection rate of biliary brush cytology remains low. Our high panc CA prevalence (63%) may partially account for this. However, the yield of two cytology brushes is greater than one. Biliary stricture dilation prior to performing brush cytology does not appear to increase diagnostic accuracy. A larger series is needed to determine the benefits (if any) from stricture dilation. Studies comparing multiple brushings with and without stricture dilation would be of interest.