David Condon

Sa1217 Are There Predictive Factors for P2 Findings on Inpatient Video Capsule Endoscopy

G A A b st ra ct s of cancers of both pancreas and biliary tract, DWI showed excellent sensitivity but slightly lower specificity than PET/CT. 4) Hyperintense region by DWI prior to treatment in all the patients with AIP completely disappeared after long term treatment. 5) Hyperintense region by DWI markedly improved 2 weeks after steroid pulse therapy, making the diagnosis of AIP possible on this ground. CONCLUSON: DWI was as good as PET/CT in assessing gallbladder and pancreatic cancer. DWI was not as useful as PET/CT in detecting bile duct and Ampulla of Vater cancer, making the additional use of MRCP imaging necessary. In the dignosis of lymph node metastases DWI appears to be more sensitive than PET/CT. DWI was useful in the assessment of curative effect of steroid therapy and also in early diagnosis utilizing the response to steroid pulse therapy in AIP patients.

Utility of Diphenhydramine for Conscious Sedation During Colonoscopy

Utility of Diphenhydramine for Conscious Sedation During Colonoscopy Wichit Srikureja, Yoshi Mineyama, John D. McCracken, David Condon Purpose: With the recent black box warning of life-threatening arrhythmia on droperidol, there is a need to look for a safer adjunct medication for sedation. We assessed the utility of intravenous diphenhydramine as an adjunct pre-medication in routine colonoscopy in a prospective, randomized, double-blinded, and placebocontrolled trial. Methods: Consecutive patients referred for colonoscopy were randomized to receive either 50mg diphenhydramine or saline placebo. The study medicationwasgiven intravenously and immediately followedby25mgofmeperidine and1mgofmidazolam.After threeminutes, patient receivedadditionalmeperidineor midazolam at the discretion of the endoscopist. The procedure could then be started. Procedure and recovery times, number of interruptions, total medication dosages, andadverse eventswererecorded.Theendoscopist and thesedatingnurse assessed the quality of the sedation on a 10-point scale (1=poor, 10=excellent). Upon discharge and at twenty-four hour post procedure, the patients assessed the quality of sedation quantitatively using the above scale and qualitatively through questionnaires. Results:A total of88patientswere randomizedwith equal number ineachgroup.The mean age was 60 6 11.5 years. Fifty-two patients (59.1%) were male. Table 1 summarizes thecomparisonofmeanvariables in eachgroup.Atdischarge,patients in the diphenhydramine group were more likely to feel drowsy (p=0.053), less likely to be back to usual self (p=0.042), and less likely to recollect the procedure (p=0.026). These differences were not seen at twenty-four hour.No adverse events were noted in either group. Conclusion: Although diphenhydramine is well tolerated with additional amnesic effect, it does not increase the quality of sedation, shorten the procedure time, or reduce thestandard sedativesdosageswhenusedasanadjunctpremedication in patients undergoing routine colonoscopy. Its role in patients difficult to sedate or in prolonged procedure such as ERCP remained to be confirmed. **198 Complications of Endoscopic Ultrasound-Guided Fine Needle Aspiration (EUS-FNA) of Pancreatic Cysts Linda S. Lee, John Saltzman, William R. Brugge, Christopher C. Thompson Purpose: EUS-FNA is generally accepted as a safe endoscopic procedure. However, data regarding complications after EUS-FNA of pancreatic cysts is limited and conflicting. The aim of this study is to ascertain the complication rates of EUS-FNA of pancreatic cysts and related risk factors. Methods: EUS-FNA of 684 pancreatic cysts at two academic institutions between 3/1996 and 10/2003were retrospectively reviewed. Following EUS-FNA, patients were tracked using discharge summaries, clinic logs, laboratory and radiologic data in the computerized hospital systems. The following data were gathered: cyst diagnosis, size, septations, degree of aspiration, use of prophylactic antibiotics, complications, and performance of ERCP on the same date. For complications, further data were collected: symptoms, temperature, hospitalization, length of hospital stay (LOS), laboratory data, and treatment during hospitalization. Results: Fifteen nonfatal complications were identified (2%, 15/684): 6 pancreatitis, 2 selflimited retroperitoneal bleeds, 2 infections, and 4 abdominal pain. Fourteen required hospitalization with an average LOS 4.6 days (range 1-14). The following predictors were examined but were not statistically significant (patients with complications vs. without complications): diagnosis (73% vs. 62% cystic neoplasm, 7% vs. 10% pseudocyst, 7% vs. 14% IPMT, 0% vs. 14% cancer), size (30.9 mm vs. 25.8 mm), septations (67% vs. 49% unilocular, 33% vs. 38% multilocular, 0% vs. 13% solid-cystic), same day ERCP (13% vs. 8%), and use of prophylactic antibiotics (100% vs. 89%). There was no increased rate of pancreatitis in patients undergoing both EUS-FNA and ERCP (n=56) (1.8%) on the same day versus only EUS-FNA (n=628) (0.8%). In a subset analysis of 76 EUS-FNAof pancreatic cysts not given prophylactic antibiotics, no complications were identified. Average cyst size was significantly smaller than the 608 cysts aspirated with prophylactic antibiotics (19.8 mm vs. 26.8 mm, p=0.003). Conclusions: In this largest reported experience of EUS-guided pancreatic cyst aspiration, the complication rate is low (2%). Small pancreatic cysts were safely aspirated without the use of prophylactic antibiotics. Cyst diagnosis, size, septations, mass, and performance of same day ERCP do not appear to be risk factors for the development of complications.

Acute food impaction: a rare presentation and complication of heterotopic gastric mucosa in the proximal esophagus

was absent, and esophageal and gastric varices were excluded by endoscopy (EGD). Due to significant impairment in quality of life and increased risk of medical complications from persistent HE, the patient and his wife were consented for an experimental procedure to reduce flow in his splenorenal shunt. Procedure: The left renal vein and splenorenal conduit were accessed via percutaneous catheter. A Simon-Nitinol Filter was deployed into the shunt with the tip directed toward the renal vein. Six 10-12mm metal coils were deployed into the filter. Angiographic flow rate was subsequently reduced. Final hepatic vein pressure gradient remained unchanged at 28mmHg at the end of the procedure. Shunt flow could not be visualized on contrast ultrasound the following day. Two weeks later, ultrasound revealed improved hepatopetal flow within the portal vein and no ascites, and EGD revealed trace esophageal varices without bleeding. HE has improved without requirement for admission while on maintenance oral lactulose. Discussion: Spontaneous portosystemic shunting should be suspected in patients with cirrhosis, persistent HE, and lack of significant ascites or esophagogastric varices. Persistent HE refractory to medical therapy can be safely controlled by percutaneous embolization of portosystemic shunts.