A comparative evaluation of early stent occlusion among biliary conventional versus wing stents.

BACKGROUND AND STUDY AIMS: Conventional plastic stents with a lumen typically have limited patency. The lumenless wing stent was engineered to overcome this problem. The objective of this study was to compare the incidence of early stent occlusion (symptomatic occlusion/cholangitis necessitating re-insertion within 90 days) for wing stents and conventional plastic stents. PATIENTS AND METHODS: Patients with biliary pathology treated with plastic biliary stenting during the period 2003-2009 comprised the study cohort. Patients who had at least one biliary wing stent placed comprised the wing stent group, whereas patients who underwent only conventional stent plastic placement comprised the conventional stent group. Patients were stratified by indication: benign biliary strictures (group 1), malignant biliary strictures (group 2), or benign biliary non-stricture pathology (group 3). The association of stent type with the occurrence of primary outcome by indication was analyzed by use of multivariable logistic regression. RESULTS: Three-hundred and forty-six patients underwent 612 ERCP procedures with placement of plastic biliary stent(s). On multivariate analysis, early stent occlusion did not differ between the wing and conventional groups in groups 1, 2, and 3. Among patients who achieved primary outcome in group 2, significantly fewer patients in the wing group had cholangitis (6.7% vs. 39.1%, P = 0.03). Among patients who achieved primary outcome in group 3, significantly fewer patients in the wing group had cholangitis (10% vs. 50%, P = 0.03). CONCLUSIONS: Early stent occlusion was similar for wing stents and conventional plastic stents. Wing stents, however, were associated with a lower incidence of cholangitis in patients with malignant biliary obstruction and benign non-stricturing biliary pathology.

Zymogen Protein C to prevent clotting without bleeding during invasive medical procedures.

Thrombophilic disorders that predispose patients to develop blood clots can be life-threatening and result in a large economic burden on healthcare expenditures. Venous Thromboembolism(VTE) (deep vein thrombosis and pulmonary embolism) are the third leading cause of death in the United States. Protein C deficiency is a common thrombophilic condition that affects an estimated 1 in 400 Americans. Zymogen Protein C (ZPC) is the precursor to Activated Protein C (APC), a pivotal endogenous anticoagulant in human blood. Patients with protein C deficiency who have roughly half the normal level of protein C are estimated to be at 10-fold increased risk of VTE. We describe the use of protein C concentrate (Ceprotin®, Baxter, Deerfield, IL) in a patient with protein C deficiency and with a previous pulmonary embolism who developed a life-threatening gastrointestinal bleed after polypectomy. The patient is a 75-year-old male at very high risk for deep vein thrombosis and possible lung emboli. He has heterozygous Protein C deficiency (50%) and heterozygosity for the prothrombin gene G20210A mutation. During a routine colonoscopy, a large 3 cm cecal polyp was identified and resected. Eight days post-procedure while performing abdominal exercise he developed a life-threatening GI bleed originating from the polypectomy site as his warfarin was becoming therapeutic on a Low Molecular Weight Heparin (LMWH) periprocedural bridge. The patient's warfarin was reversed with vitamin K, and LMWH and warfarin were discontinued. To prevent thrombosis, he was started on ZPC until anticoagulation could be safely restarted. During endoscopy, the bleeding site was treated with an injection of 1:10,000 dilution of epinephrine, followed by cauterization and placement of endoclips (4 metal staples). Three days after endoscopic repair LMWH was restarted with warfarin. Sixteen months post-bleed, the patient remains on life-long warfarin without further episodes of bleeding or thrombosis. Zymogen Protein C concentrate (Ceprotin®, Baxter Deerfield, IL) should be strongly considered for peri-procedural management of any patient with protein C deficiency and previous thromboembolism.

Autoimmune pancreatitis: current diagnostic criteria are suboptimal.

BACKGROUND AND AIMS: The preoperative diagnosis of autoimmune pancreatitis (AIP) is difficult, given its similar clinical presentation to pancreatic cancer. The aims of the study are to describe our center's experience with AIP and apply the Japanese AIP diagnostic criteria to a cohort of patients with histologically-proven AIP in order to assess their performance characteristics. METHODS: A prospective pathology database was queried for AIP patients who were evaluated and/or treated at Johns Hopkins Hospital from 2002 to 2009. AIP histology was defined by the presence of lymphoplasmacytic infiltration, periductal inflammation, fibrosis, and periphlebitis. Imaging, clinical, and biochemical data were analyzed. RESULTS: Thirty patients had pancreatic resection with pathological confirmation of AIP. Imaging revealed pancreatic mass (45%), focal prominence without mass lesion (24%), diffuse enlargement (17%), and normal pancreas (14%). Twenty-four patients underwent an endoscopic retrograde cholangiopancreatography and/or magnetic resonance cholangiopancreatography, and 4/24 (17%) had pancreatic ductal narrowing or irregularity. Extrapancreaticobiliary organ involvement was found in 6% (n = 2) of patients. Biliary strictures were present in 87% of patients. Of 16 patients who underwent preoperative tissue biopsy, 10 had non-diagnostic pathology, five had cellular atypia, and one had AIP. Serum immunoglobulin G4 (IgG4) levels were elevated in 12 of 29 (41%) patients. Three (10%) patients had evidence of extrapancreatic manifestations of AIP. When applying the Japanese criteria to the 27 patients who had serum IgG4 measurement, preoperative biopsy, and cross-sectional abdominal imaging, only 44% of the patients would have been diagnosed accurately. CONCLUSIONS: When applied to a highly-selected single-center referral population in the USA, current Japanese guidelines for the diagnosis of AIP are found to have suboptimal sensitivity.

Hospital volume as a predictor for undergoing cholecystectomy after admission for acute biliary pancreatitis.

OBJECTIVE: We explored whether admission volumes for cholecystectomy (CCY) and pancreatitis were associated with receiving CCY after hospitalization for acute biliary pancreatitis (ABP). METHODS: We identified admissions for ABP in the Nationwide Inpatient Sample between 1998 and 2003. We used multivariate analysis to assess the association between likelihood of CCY and hospital volumes of CCY, pancreatitis, and endoscopic retrograde cholangiopancreatography (ERCP). RESULTS: The overall rate of CCY for ABP was 50%. After adjustment for confounders, the likelihood of CCY increased with every quartile of CCY volume relative to the bottom quartile (adjusted odds ratios of 4.36, 7.92, and 12.51 for quartiles 2, 3, and 4, respectively, P < 0.0001). Pancreatitis volume was inversely correlated with likelihood of CCY (adjusted odds ratios of 0.72, 0.62, and 0.48 for quartiles 2, 3, and 4, respectively, vs bottom quartile, P < 0.01). Admissions to hospitals in the top quartile for ERCP volume (>35 ERCPs/yr) had 15% lower odds of CCY than the lowest quartile. Patients from rural areas and with lower income were disproportionately admitted to hospitals with lower CCY volumes. CONCLUSIONS: US hospitals are not achieving targets for CCY after ABP as set by national and international guidelines. Centers with smaller CCY volumes are the least adherent to recommendations for CCY possibly because of hospital-level resource limitations.

Are large side holes associated with reduced rates of pancreatic stent occlusion? Results of a prospective study.

CONTEXT: Pancreatic stents are frequently clogged at the time of removal. There is limited data regarding the factors associated with stent occlusion. OBJECTIVES: To estimate the frequency of stent occlusion at the time of removal, to study the accuracy of endoscopic prediction of occlusion, and to determine the factors associated with clogged pancreatic stents. SETTING: Consecutive patients at 4 academic medical centers undergoing removal of a previously placed pancreatic stent were prospectively enrolled. PATIENTS: A total of 68 patients were enrolled between August 2007 and July 2008. INTERVENTIONS: Following removal, stent occlusion was immediately assessed by complete lack of water flow from the duodenal end and side holes of the stent. MAIN OUTCOME MEASURE: Survival analysis was performed using a Kaplan-Meier and Cox Regression model. RESULTS: Indications for stent placement included chronic pancreatitis (n=23), pancreatic duct leak (n=7), prevention of post-ERCP pancreatitis (n=28), pseudocyst drainage (n=1), pancreas divisum (n=8), and pancreatic duct stricture without chronic pancreatitis (n=1). Standard Geenen (Cook Endoscopy, Winston-Salem, NC, USA) pancreatic stents were placed in 53 patients (77.9%). The majority of stents (42, 61.8%) were completely occluded at the time of removal. Median time to stent occlusion was 35 days (95% CI: 30-40 days). Stent type, diameter, length, number of small side holes, and indication for placement were not predictive of subsequent stent occlusion. However, stents with at least 4 large, flange-associated side holes were 54% less likely to be clogged upon removal (HR=0.46, P=0.029). CONCLUSIONS: Pancreatic stents are mostly occluded as early as 1 month after insertion. Larger side holes may prevent stents within the pancreas from becoming clogged.

Patient- and cyst-related factors for improved prediction of malignancy within cystic lesions of the pancreas.

BACKGROUND AND AIMS: Early diagnosis of cancer in pancreatic cysts is important for timely referral to surgery. The aim of this study was to develop a predictive model for pancreatic cyst malignancy to improve patient selection for surgical resection. METHODS: We performed retrospective analyses of endoscopic ultrasound (EUS) and pathology databases identifying pancreatic cysts with available final pathological diagnoses. Main-duct intraductal papillary mucinous neoplasms (IPMNs) were excluded due to the clear indication for surgery. Patient demographics and symptoms, cyst morphology, and cyst fluid characteristics were studied as candidate risk factors for malignancy. RESULTS: 270 patients with pancreatic cysts were identified and analyzed (41% men, mean age 61.8 years). Final pathological diagnoses were branch-duct IPMN (n = 118, 50% malignant), serous cystadenoma (n = 71), pseudocyst (n = 37), mucinous cyst adenoma/adenocarcinoma (n = 36), islet cell tumor (n = 4), simple cyst (n = 3), and ductal adenocarcinoma with cystic degeneration (n = 1). Optimal cut-off points for surgical resection were cyst fluid carcinoembryonic antigen (CEA) > or =3,594 ng/ml, age >50, and cyst size >1.5 cm. Cyst malignancy was independently associated with white race (OR = 4.1, p = 0.002), weight loss (OR = 3.9, p = 0.001), cyst size >1.5 cm (OR = 2.4, p = 0.012), and high CEA > or =3,594 (OR = 5.3, p = 0.04). In white patients >50 years old presenting with weight loss and cyst size >1.5 cm, the likelihood of malignancy was nearly sixfold greater than in those patients who had none of these factors (OR = 5.8, 95% CI = 2.1-16.1, p = 0.004). CONCLUSIONS: Risk factors other than cyst size are important for determination of malignancy in pancreatic cysts. Exceptionally high cyst fluid CEA levels and certain patient-related factors may help to better predict cyst malignancy and the need for surgical treatment.

Delayed release pancrelipase for treatment of pancreatic exocrine insufficiency associated with chronic pancreatitis.

Pancreatic enzyme supplements (PES) are used in chronic pancreatitis (CP) for correction of pancreatic exocrine insufficiency (PEI) as well as pain and malnutrition. The use of porcine pancreatic enzymes for the correction of exocrine insufficiency is governed by the pathophysiology of the disease as well as pharmacologic properties of PES. Variability in bioequivalence of PES has been noted on in vitro and in vivo testing and has been attributed to the differences in enteric coating and the degree of micro-encapsulation. As a step towards standardizing pancreatic enzyme preparations, the Food and Drug Administration now requires the manufacturers of PES to obtain approval of marketed formulations by April 2010. In patients with treatment failure, apart from evaluating drug and dietary interactions and compliance, physicians should keep in mind that patients may benefit from switching to a different formulation. The choice of PES (enteric coated versus non-enteric coated) and the need for acid suppression should be individualized. There is no current standard test for evaluating adequacy of therapy in CP patients and studies have shown that optimization of therapy based on symptoms may be inadequate. Goals of therapy based on overall patient presentation and specific laboratory tests rather than mere correction of steatorrhea are needed.

Disparities in demographics among patients with pancreatitis-related mortality.

CONTEXT: Hospital admissions for pancreatitis are increasing. Factors involved in inpatient mortality have not been previously assessed on a large-scale basis. OBJECTIVE: The aim was to study factors associated with pancreatitis-related death in hospitalized patients. SETTING: Retrospective analysis of the 2004 U.S. Healthcare Cost and Utilization Project (HCUP) database was performed using "pancreatitis" as admitting diagnosis and "mortality" as primary endpoint. MAIN OUTCOME MEASURES: Age, race, gender, income, length of stay, number of diagnoses, and number of procedures were identified as candidate risk factors associated with death. DESIGN: Univariate and multivariate logistic regression analyses were performed to identify significant covariates. RESULTS: In 2004, total of 78,864 patients were admitted with pancreatitis; 2,129 (2.7%) patients died. Complete data were available for 57,068 patients. Age greater than 65 was 3 times more often associated with mortality (OR=2.92; P<0.001), while females were 19% less likely to die (OR=0.81; P<0.001). African American patients were 18% more likely to die than whites (OR=1.18, P=0.025), and increasing length of stay was associated with increasing mortality (more than 14 days compared with less than 3 days: OR=1.24; P=0.004). Patients with more than 3 diagnoses and more than one hospital procedure were 17 times (OR=16.7; P<0.001) and 5 times (OR=5.42; P<0.001) more likely to die, respectively. Compared to the lowest income quartile, patients in the 2nd and 3rd quartiles were 19% (OR=0.81; P=0.004) and 17% (OR=0.83; P=0.016) less likely to die, respectively. CONCLUSION: Age greater than 65 years, male gender, multiple diagnoses, African American race and low income are strongly associated with inpatient mortality from pancreatitis. Increased number of procedures and longer length of stay are also highly correlative with death.

Awareness of guidelines and trends in the management of suspected pancreatic cystic neoplasms: survey results among general gastroenterologists and EUS specialists.

BACKGROUND: Although pancreatic cystic neoplasms are widely recognized, practice habits among physicians and awareness of consensus guidelines are currently unknown. OBJECTIVES: To assess the awareness of guidelines and describe variability in practice habits among 2 groups: (1) "general group" of gastroenterologists and surgeons and (2) "EUS group" of specialists in EUS. DESIGN: An online survey was sent to randomly selected gastroenterologists and surgeons and e-mailed to members of the American Society for Gastrointestinal Endoscopy (ASGE) Special Interest Group in EUS (EUS-SIG). RESULTS: Response rate for the general group was 8.8% (220/2500) and 9.7% for the EUS group (42/431). EUS specialists were mostly in academic practice (66.7% vs 36.3%, P < .001) and reported seeing 21 to 50 cysts per year (54.8% vs 12.3%, P < .001). The majority of the general group (64.1%) was unaware of any published practice guidelines, compared with 33.3% of EUS specialists (P < .001). Awareness of ASGE guidelines was more frequently reported than other guidelines in both groups and yet was still <50% for each group. Both demonstrated moderate consistency with the International Association of Pancreatology guidelines, appropriately answering 66.7% of the questions. For 9-mm lesions, only 25% of the questions were correctly answered in each group. EUS specialists were less likely to refer main-duct intraductal papillary mucinous neoplasms (IPMN) for surgery and more likely to opt for EUS-guided FNA (compared with high-resolution CT, MRCP, or surgery) for 9-mm, 22-mm, and 34-mm branch-duct IPMNs (P

Racial disparities in cholecystectomy rates during hospitalizations for acute gallstone pancreatitis: a national survey.

BACKGROUND: Practice guidelines advocate performing cholecystectomy for acute gallstone pancreatitis during the same hospitalization stay. Our objectives were to determine nationwide rates of adherence to these guidelines in the United States and whether this varied with race and ethnicity. METHODS: We queried the Nationwide Inpatient Sample (NIS) to identify admissions for acute gallstone pancreatitis between 1998 and 2003. We calculated overall and race-specific proportions of patients who underwent cholecystectomy or endoscopic retrograde cholangiopancreatography (ERCP) prior to discharge. We used multivariate analysis to determine racial effects while adjusting for age, comorbidity, health insurance payer, and hospital factors. RESULTS: The overall rate of cholecystectomy was 51% and that of either cholecystectomy or ERCP was 62%. Cholecystectomy rates were lower among African Americans (AAs) and Asians compared to Whites (44% and 43%, respectively, vs 50%, P < 0.001). After multivariate adjustment, the odds of cholecystectomy was lower in AAs (OR 0.68, 95% CI 0.63-0.73) and Asians/Pacific Islanders (OR 0.75, 95% CI 0.65-0.87) relative to Whites, while rates were modestly higher among Hispanics (OR 1.12, 95% CI 1.03-1.22). AAs were less likely to receive ERCP than Whites (OR 0.71, 95% CI 0.65-0.78). In contrast, Asians/Pacific Islanders (OR 1.40, 95% CI 1.16-1.69) and Hispanics (OR 1.19, 95% CI 1.09-1.29) were more likely to receive ERCP than Whites. CONCLUSIONS: Despite practice guidelines, about only half of admissions for gallstone pancreatitis receive cholecystectomy during the same hospitalization, and cholecystectomy rates vary substantially by race. These findings raise concerns regarding suboptimal healthcare delivery.

Etanercept, a TNF-alpha binding agent, is ineffective in the prevention of post-ERCP pancreatitis in canines.

CONTEXT: The incidence of post-ERCP pancreatitis is 1-22%. It continues to be a difficult problem for endoscopist and patient. Uncovering an agent that may be used to prevent its occurrence is critical. OBJECTIVE: The aim of our study was to investigate the role of etanercept in the prevention of post-ERCP pancreatitis. DESIGN: Endoscopic retrograde pancreatography (ERP)-induced injury was performed in dogs using a previously established endoscopic model of post-ERCP pancreatitis. ANIMALS: Eight study dogs underwent ERP: 4 were pre-treated with etanercept one day before the procedure and 4 were untreated. In addition, three control dogs not undergoing ERP were also studied. MAIN OUTCOME MEASURES: Serum levels of amylase, lipase, and TNF-alpha, as well as the ratio of urinary trypsinogen activation peptide (TAP) and urinary creatinine, were measured before and after ERP. Necropsy was performed on post-operative day 5. All pancreatic specimens were graded by two blinded pathologists according to a validated scoring system. RESULTS: Eight study dogs developed mild to moderate clinical pancreatitis with hyperamylasemia (11,538+/-4,065 U/L vs. 701+/-157 U/L; post-ERP peak levels vs. baseline values: P<0.001) and hyperlipasemia (3,637+/-2,333 U/L vs. 246+/-125 U/L; P=0.003). Mean total injury score was significantly elevated in study dogs compared to control dogs (6.16+/-1.85 vs. 1.06+/-0.49; P=0.001). There were escalating total injury scores concordant with more elaborate methods of endoscopically-induced injury although the trend did not reach the statistical significance (P=0.223). When comparing untreated to etanercept-treated dogs, there were no significant differences in serum amylase levels (P=0.903), serum lipase levels (P=0.771), TAP/creatinine urinary ratio (P=0.912), and pancreatic injury score (P=0.324). CONCLUSION: Etanercept is ineffective in prevention of mild to moderate post-ERCP pancreatitis in canines. ERP-induced pancreatic injury can be used as a reliable animal model for studies investigating therapy and prevention of post-ERCP pancreatitis.

EUS-guided submucosal implantation of a radiopaque marker: a simple and effective procedure to facilitate subsequent surgical and radiation therapy.

BACKGROUND: Endosonography (EUS) is widely used for locoregional staging of malignant GI tumors. Delineation of a tumor's margins with a long-lasting fluoroscopically visible material will facilitate subsequent surgical and radiation therapy. OBJECTIVE: To assess the feasibility of EUS-guided submucosal implantation of a radiopaque marker in a porcine model. SETTING: Survival experiments on four 50-kg pigs. METHODS: A linear array echoendoscope was introduced into the esophagus and advanced to the stomach. With a 19-gauge FNA needle, a submucosal bleb was created by injecting 3 mL of normal saline solution into the gastric and esophageal wall followed by injection of 1 mL of tantalum suspension under fluoroscopic observation. Fluoroscopy was repeated after 1, 2, and 4 weeks followed by euthanasia and necropsy. MAIN OUTCOME MEASUREMENTS: Long-term depositions of the marker in the injection sites. RESULTS: Submucosal injections of tantalum were easily performed through the 19-gauge FNA needle, resulting in good fluoroscopic opacification of injected material. Follow-up fluoroscopy in 1, 2, and 4 weeks demonstrated stable deposition of the tantalum at the sites of injection. There were no complications during and after the tantalum implantation. Histologic examination of the injection sites demonstrated submucosal tantalum depositions without signs of infection, inflammation, tissue damage, or necrosis. LIMITATIONS: Animal experiments with 4 weeks' follow-up. CONCLUSIONS: EUS-guided implantation of tantalum as a radiopaque marker into the submucosal layer of the GI tract in a porcine model is technically feasible and safe. Long-lasting fluoroscopically visible tantalum markings could facilitate subsequent surgical and radiation therapy.

How good is capsule endoscopy for detection of periampullary lesions? Results of a tertiary-referral center.

BACKGROUND: Ampullary adenomas are increasingly being recognized, particularly in patients with familial adenomatous polyposis. A capsule endoscopy (CE) is routinely recommended for surveillance of small-intestinal polyposis. Performance characteristics of CE for the detection of periampullary lesions are unclear. OBJECTIVE: To evaluate the ability of CE to detect the major duodenal papilla. DESIGN AND PATIENTS: A total of 146 consecutive CE studies were reviewed by 2 CE gastroenterologists at 5 frames per second. Primary outcome was visualization of the major duodenal papilla. Discrepancies were reviewed by 5 CE gastroenterologists. SETTING: A tertiary-referral center. MAIN OUTCOME MEASUREMENTS: The ability of CE to detect the duodenal papilla. RESULTS: Among 146 consecutive CE studies, 21 were excluded: capsule retention (3), patient age <18 years (6), duplicate study (8), and prior surgery disrupting duodenal anatomy (4). Of the remaining 125 studies, indications were the following: obscure GI bleeding (45.6%), iron deficiency anemia (19.2%), abdominal pain (17.6%), diarrhea (10.4%), and Crohn's disease (4.8%). In total, 13 major duodenal papillae were visualized. The median time of detection was 31 seconds after the first duodenal image. This translates to a CE sensitivity of 10.4% for detection of the major papilla. LIMITATION: Papilla position was not verified by an EGD. CONCLUSIONS: CE has limited sensitivity to visualize the major papilla and lesions in the periampullary small intestine. Nondiagnostic CE studies must not be relied upon as proof that small-bowel lesions do not exist. Consideration should be given for an enteroscopy or side-viewing duodenoscopy in cases where significant clinical concern exists for unrecognized periampullary lesions. The current recommendations about surveillance for small-bowel polyposis should be revised.

Comparative study of endoscopic clips: duration of attachment at the site of clip application.

BACKGROUND: Several designs of endoscopic clips are now commercially available, and the indications for endoclip application are rapidly expanding. However, very limited data have been published to aid in choosing between the different types of endoclips. OBJECTIVE: To compare the duration of clip attachment between all commercially available endoclips. SETTING: Long-term experiments on 50-kg pigs under general anesthesia. DESIGN AND INTERVENTIONS: Upper endoscope was inserted into the stomach. One clip of each type (Resolution clip, TriClip, and HX-5L clip) was placed along the same gastric fold at a distance of 0.5 to 1 cm from each other. The animals were recovered. In pig nos. 1 and 2, repeat endoscopy was performed after 2 and 4 weeks. In pig nos. 3 to 5, endoscopy was repeated after 1, 2, and 5 weeks. MAIN OUTCOME MEASUREMENTS: Duration of clip retention at the site of application. RESULTS: In all animals, only the Resolution endoclip remained attached to the site of application for the entire duration of the study (4-5 weeks). No TriClips or HX-5L clips were attached at the 4- to 5-week follow-up endoscopies. Most of the TriClips (67%) detached within the first week after application. Most of the HX-5L clips (80%) dislodged within the first 2 weeks of follow-up. LIMITATIONS: The study was performed in a porcine model with a small number of animals. CONCLUSIONS: The Resolution clip has the longest duration of retention at the site of application (more than 4-5 weeks) and should be preferred when long-term attachment of endoclips is necessary.

EUS-guided portal vein carbon dioxide angiography: a pilot study in a porcine model.

BACKGROUND: Direct portal vein (PV) angiography can provide valuable clinical information but is not performed due to the high risk of complications. OBJECTIVE: To assess the feasibility of EUS-guided PV angiography with a small-caliber FNA needle by using carbon dioxide (CO(2)) as a contrast agent in a porcine model. SETTING: Acute experiments with 50-kg pigs under general anesthesia. DESIGN AND INTERVENTIONS: Under linear array EUS guidance, the intrahepatic PV branch was punctured with a 25-gauge FNA needle. Portal venography was performed with iodinated contrast (Hypaque) and then with medical grade CO(2). After portography, the needle was removed from the PV and the animals were observed for 30 minutes, then euthanized for necropsy. MAIN OUTCOME MEASUREMENT: Ability to visualize portal anatomy. RESULTS: Six animal experiments were performed without complications. EUS-guided PV puncture with 25-gauge FNA needle was technically straightforward. Injection of ionic iodinated contrast through the 25-gauge FNA needle was arduous (mean [+/-SD] pressure 76.7 +/- 5.2 pounds per square inch [psi]), resulting in short (6.02 +/- 1.15 seconds) and poor opacification of the PV (visualization score 1.33 +/- 0.52). CO(2) injection through a 25-gauge needle was simple and easy (pressure 20.8 +/- 2.0 psi), producing prolonged (19.83 +/- 1.68 seconds) opacification of the entire portal system (visualization score 4.33 +/- 0.52). There was a statistically significant difference in all compared parameters (P < .0001) favoring injection of CO(2) over viscous iodinated contrast during portal angiography through a 25-gauge FNA needle. Postmortem examination revealed no active bleeding and no damage to the liver, other intra-abdominal organs, or blood vessels. LIMITATION: Acute animal experiments. CONCLUSIONS: EUS-guided portal venography with CO(2) using a small (25 gauge) FNA needle appears feasible, technically simple, and safe.