240 The role of ERCP in the management of pancreatic pseudocysts (PP): further support for an ERCP-based algorithm Mergener Klaus, Mundorf Jeffrey B, Pappas Theodore N, Robuck- Mangum Gail, Jowell. Paul S, Branch Malcolm S, Swartz Katrina L, Baillie* John. Duke University Medical Center, Durham, NC, United States. Purpose: Our previously published algorithm (1) for the management of PP assumes a high recurrence rate after percutaneous drainage when there is pancreatic duct (PD) obstruction or the pseudocyst communicates with the duct. We have recommended surgical intervention in this group. In an earlier review (2), this triage of PP cases seemed to offer good results. In order to validate the algorithm, we applied it to a large number of patients with PPs seen at our tertiary care center. Gastroenterologists and surgeons managing PP patients were allowed to make independent (i.e. not algo- rithm-driven) decisions regarding management. Methods: Computerized endoscopic, surgical and radiological databases were reviewed retrospectively to identify patients undergoing percutaneous and surgical treatment of PPs (as defined by Atlanta criteria) between 1992 and 1998. Patients with multiple (.3) PPs, those suspicious for malig- nancy, and those treated endoscopically were excluded. Health status, comorbidities, severity of pancreatitis, PP characteristics, PD anatomy, management strategy and outcome (success, recurrence, complications) were recorded. Results: 121 Patients (ages 25–79 yr) were included in this analysis. 77 (64%) patients followed the algorithm and 44 (36%) did not follow the algorithm. Overall, 60/77 (78%) interventions that followed the algorithm succeeded. This is compared to 18/44 (41%) that succeeded but did not follow the algorithm (p , 0.02). Patients were well matched for age, health status, severity of pancreatitis, size, number and location of PPs. Conclusions: ERCP is a valuable tool for predicting the outcome of percutaneous drainage of PPs. In the presence of PD obstruction or PD/PP communication, surgery is likely to be needed. The role of endoscopic management (e.g. cystgastrotomy) was not addressed by this study, nor was the contribution of new imaging techniques (e.g. MRCP). References: (1) Ahearne PM et al., Am J Surg 1992;163:111–116, (2) Mundorf JB et al., Grastrointest Endosc 1996;43:411A. 241 Pancreatitis-associated protein (PAP) levels in pancreatic juice in patients with pancreatic diseases Motoo MD FACG Y, Watnabe MD H, Sawabu MD N. Dept of Internal Med. Cancer Res Instit. Kanazawa Univ. Japan. Purpose: PAP is barely expressed in the normal pancreas, but is overex- pressed in the acinar cells in acute pancreatitis. We have reported an ectopic expression of PAP in cancer cells and serum PAP levels in digestive diseases. In this study, we aimed to determine PAP levels in pancreatic juice (PJ) in patients with pancreatic diseases. Methods: PJ was collected by endoscopic aspiration with selective can- nulation after secretin stimulation from 22 patients with pancreatic cancer (PC), 49 patients with chronic pancreatitis (CP) and 15 control patients. PJ-PAP and serum (S-) PAP levels were determined using an ELISA kit (Dynabio, La Gaude, France). Results: PJ-PAP was positive (.350 ng ml) in 55% of PC and 25% of CP. PJ-PAP levels of PC (2333 6 3289 ng ml, mean 6 SD) were significantly (P , 0.002) higher than those of CP (579 6 962 ng ml). PJ-PAP levels in CP were not significantly higher than those in controls (142 6 66 ng ml). PJ-PAP was positive in 6 of 18 S-PAP-negative cases among 28 patients with PC or CP in whom PJ-PAP and S-PAP were both measured. Four of the 6 PJ-PAP-positive, S-PAP-negative patients had PC. Conclusions: PJ-PAP levels in PC were higher than those in CP, reflecting an ectopic expression of PAP in cancer cells. 242 Is outpatient therapeutic ERCP practical? Naik Arun C, Kasmin* Franklin E, Cohen Seth A, Siegel Jerome H. Beth Israel Medical Center, North Division, New York, NY. Purpose: To assess the need for parenteral pain management in the absence of complications in a large group of consecutive patients undergoing therapeutic ERCP. Methods: We prospectively analysed all therapeutic ERCP over a 7 month period performed by a referral center. Indications for the admission of pts after ERCP included sphincterotomy, difficult cannulation, the need for IV antibiotics, or pain after the procedure. Pts who underwent uncomplicated therapy (i.e. stent change) were discharged following the procedure. Inpatients were offered parenteral meperidine or oral oxycodone/APAP as needed. Ambulatory pts were given oral oxycodone/APAP. All pts were followed for complications, and the use of narcotic pain medication was assessed. Outpatients who used oral narcotics were asked whether they would have preferred parenteral narcotics. The use of pain medication was analyzed according to indication for the ERCP, the intervention carried out, and inpatient/ambulatory status. Patients who developed complications were excluded from analysis. We sought to assess the necessity of inpatient admission for parenteral pain management according to indications for ERCP and interventions carried out. Results: We studied 305 pts. 185 were female, mean age–55.8 years. Indications were: cbd stones and benign strictures–22%, acute recurrent pancreatitis–20%, biliary dyskinesia/abd. pain–20%, chronic pancreatitis– 13%, malignant cbd obstruction–9%, other–16%. 198 pts were admitted after ERCP and 107 were outpatients. Of the 305 pts, 198 were admitted. 53 of these pts developed compli- cations and were excluded. Of the 145 inpatients without complications, 41 (28%) required parenteral pain medications. 20 additional pts. (13.8%) used oral narcotic pain meds. Parenteral narcotic use according to indica- tion for ERCP was: Biliary dyskinesia/Abd pain–17 of 44 (39%), recurrent pancreatitis– 8 of 46 pts (17%), malignant obstruction– 4 of 28 pts (14%), cbd stones– 4 of 64 pts (6%), chronic pancreatitis– 4 of 33 pts (12%), other– 4 of 38 (10%). Parenteral pain medication according to intervention was: Sphincterotomy–27 of 86 (31%), Precut–7 of 24 (29%), sphincter manometry–5 of 16 (31%), difficult cannulation–11 of 27 (41%), pancre- atic therapy–10 of 46 (22%). Of 107 outpatients, 34 (31.8%) required oral narcotics. 10 of the 107 (10%) would have preferred parenteral narcotics; 6 with chronic pancre- atitis, 4 with biliary dyskinesia/pain. Conclusions: A high percentage of pts who undergo therapeutic ERCP will require inpatient care for pain yet not have complications per se. 94 of 198 pts (47%) admitted after ERCP required inpatient management of pain or complications. In a referral ERCP setting, pts should undergo aggressive therapeutic ERCP on an inpatient basis. 243 Enteral infusions of elemental diets stimulate pancreatic enzyme secretion in normal subjects but not in patients with acute pancreatitis O’Keefe* Stephen, Abou-Assi Souheil G, Lee Ronzo B, Anderson Frank P. Medical College of Virginia, Richmond, VA, United States. Purpose: Conventional management of acute pancreatitis has been to restrict oral feeding and use intravenous nutrition (TPN) for fear that food may stimulate proteolytic enzyme production and exacerbate the disease process. However, recent controlled clinical trials have provided evidence that enteral feeding with elemental formulae (ED) is well tolerated, and safer and cheaper than TPN. Our purpose was to determine whether the beneficial effects of enteral elemental diets (ED) in the nutritional man- agement of acute pancreatitis (AP) are due to their ability to “rest the pancreas”. Methods: In order to determine whether EDs are as effective as TPN in resting the pancreas, enzyme secretory responses during 4 – 6 hr duodenal 2482 Abstracts AJG – Vol. 95, No. 9, 2000