The issue of whether it is possible to prevent pancreatic

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2004;2: Prophylaxis of ERCP-Related Pancreatitis: A Randomized, Controlled Trial of Somatostatin and Gabexate Mesylate ANGELO ANDRIULLI,* LUIGI SOLMI, SILVANO LOPERFIDO, PIETRO LEO, VIRGINIA FESTA, ANGELO BELMONTE, FULVIO SPIRITO,* MICHELE SILLA, # GIOVAMBATTISTA FORTE,** VITTORIO TERRUZZI, GIORGIO MARENCO, ENRICO CILIBERTO, ANTONIO SABATINO, FABIO MONICA, MARIA RITA MAGNOLIA, and FRANCESCO PERRI,* ON BEHALF OF THE ITALIAN GROUP FOR PREVENTION OF ACUTE PANCREATITIS (GIPPA STUDY 2) *Division of Gastroenterology, Casa Sollievo della Sofferenza Hospital Istituto Ricovero Cura Carattere Scientifico, San Giovanni Rotondo; Division of Gastroenterology, Sant Orsola Malpighi University Hospital, Bologna; Division of Gastroenterology, S. Maria dei Battuti General Hospital, Treviso; Division of Gastroenterology, Santo Spirito General Hospital, Rome; Division of Gastroenterology, S.S. Annunziata General Hospital, Cosenza; # Division of Gastroenterology, Civile General Hospital, Vasto; **Division of Gastroenterology, San Sebastiano General Hospital, Caserta; Division of Gastroenterology, Valduce General Hospital, Como; Division of Gastroenterology, Santa Corona General Hospital, Pietra Ligure; and Division of Gastroenterology, Civile General Hospital, Crotone, Italy Background & Aims: It still is debated whether post endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis can be prevented by administering either somatostatin or gabexate mesylate. The aim of the study is to assess the efficacy of a 6.5-hour infusion of somatostatin or gabexate mesylate in preventing ERCP-related complications. Methods: In a double-blind multicenter trial, 1127 patients undergoing ERCP were randomly assigned to intravenous administration of somatostatin (750 g; n 351), gabexate mesylate (500 mg; n 381), or placebo (saline; n 395). The drug infusion started 30 minutes before and continued for 6 hours after endoscopy. Patients were evaluated clinically, and serum amylase levels were determined at 4, 24, and 48 hours after endoscopy. Results: No significant differences in incidences of pancreatitis, hyperamylasemia, or abdominal pain were observed among the placebo (4.8%, 32.6%, and 5.3%, respectively), somatostatin (6.3%, 26.8%, and 5.1%, respectively), and gabexate mesylate groups (5.8%, 31.5%, and 6.3%, respectively). Univariate analysis of patient characteristics and endoscopic maneuvers showed that a Freeman score >1 (P < ), >3 pancreatic injections (P < ), and precut sphincterotomy (P 0.01) were significantly associated with post-ercp pancreatitis. At multiple logistic regression analysis, >3 pancreatic injections (odds ratio [OR], 1.95; 95% confidence interval [CI], ) and a Freeman score >1 (OR, 1.47; 95% CI, ) retained their predictive power. Conclusions: Long-term (6.5-hr) administration of either somatostatin or gabexate mesylate is ineffective for the prevention of post-ercp pancreatitis. Pancreatic injury seems to be related to difficulty in common bile duct access. The issue of whether it is possible to prevent pancreatic damage by the prophylactic administration of somatostatin or gabexate mesylate is still open to discussion. 1 Large-scale prospective studies reported a 4-fold reduction in acute pancreatitis compared with placebo with the prophylactic use of either gabexate 2 or somatostatin. 3 Meta-analyses of all available trials have confirmed these results. 4 Two randomized clinical trials were made available after the meta-analytical results were published. 5,6 The first trial, of patients believed to be at risk for postprocedure pancreatitis who were treated prophylactically with either somatostatin or gabexate, ended up with negative effects. 5 By including these negative results in our previous meta-analysis, the benefit of somatostatin disappeared, whereas that of gabexate persisted. 5 Because 35 patients had to be treated with a 12-hour infusion of gabexate to prevent a single episode of post endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis, 5 the economic viability of this expensive drug regimen remains questionable. The second study reconfirmed the benefit of gabexate, which was evident even after a shorter infusion time (6.50 hr) 6 ; unfortunately, because a placebo group was lacking, the last trial cannot be included in our updated metaanalysis. 5 Abbreviations used in this paper: CI, confidence interval; ERCP, endoscopic retrograde cholangiopancreatography; GIPPA, Italian Group for Prevention of Acute Pancreatitis; OR, odds ratio by the American Gastroenterological Association /04/$30.00 PII: /S (04)

2 714 ANDRIULLI ET AL. CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 8 Table 1. Baseline Characteristics Placebo Gabexate Somatostatin P Demographic characteristics Patients Sex (male/female) 220/ / / Age (yr) a Age 60 yr 273 (69.1) 255 (66.9) 250 (71.2) 0.45 Previous cholecystectomy 133 (33.7) 107 (28.4) 121 (34.8) 0.14 Main indications for ERCP Choledocolithiasis 283 (71.6) 308 (80.8) 275 (78.3) Stenosis of the main bile duct 113 (28.5) 89 (23.4) 90 (25.6) 0.25 Cholangitis 59 (14.9) 50 (13.1) 52 (14.8) 0.72 Biliary pain 209 (52.9) 205 (53.8) 188 (53.6) 0.97 Risk factors for ERCP pancreatitis History of acute pancreatitis 60 (15.2) 71 (18.6) 62 (17.7) 0.42 History of pancreatitis after ERCP 5 (1.3) 4 (1.0) 4 (1.1) 0.96 Common bile duct diameter 8 mm 123 (31.1) 130 (34.1) 128 (36.5) 0.30 Age 35 yr 24 (6.1) 30 (7.9) 21 (6.1) 0.50 Suspected sphincter of Oddi dysfunction 20 (5.0) 19 (4.9) 24 (6.8) 0.47 Difficulty of ERCP Freeman score b Grade 1, 1 5 attempts 252 (63.8) 239 (62.7) 222 (63.2) 0.81 Grade 2, 5 15 attempts 98 (24.8) 94 (24.7) 85 (24.2) Grade 3, 15 attempts 29 (7.3) 25 (6.6) 21 (6.0) Grade 4, unsuccessful 16 (4.1) 23 (6.0) 23 (6.6) Schutz score 1, simple diagnostic ERCP 55 (13.9) 55 (14.4) 56 (16.0) 2, simple therapeutic ERCP 194 (49.1) 193 (50.7) 157 (44.9) 3, complex diagnostic ERCP 3 (0.8) 2 (0.5) 8 (2.5) , complex therapeutic ERCP 105 (26.6) 93 (24.4) 99 (29.3) 5, very advanced ERCP 38 (9.6) 38 (10.0) 31 (8.9) NOTE. Values expressed as number of patients; mean 1 SD; or number (percent). a One-way analysis of variance (F test). b Number of attempts in cannulating the papilla of Vater. These contrasting data have been related to differences in schedules of administration of the 2 drugs; namely, length of infusion time 7 : a short infusion (2.5 hr) in the negative trial 5 and an 8- or 12-hour infusion in the 2 positive studies. 2,3 Evidence exists showing a time delay of several hours (median, 4.5 hr) between injury of the pancreas secondary to ERCP and time of onset of symptoms of pancreatitis 8 ; thus, a short-term infusion might not be protective. With the intent to verify whether an extended infusion, up to 6 hours after the endoscopic procedure, of either somatostatin or gabexate might benefit patients undergoing ERCP, we readdressed the value of prophylactic administration of the 2 drugs in a placebo-controlled multicenter clinical trial (Italian Group for Prevention of Acute Pancreatitis, the GIPPA 2 study). Methods Patients This double-blind, randomized, placebo-controlled clinical trial was conducted in 9 Italian hospital-affiliated centers between April 2002 and December Participating endoscopists were highly trained operators who had been performing ERCPs for a mean of 13 6 years (range, 5 21 yr). Moreover, each of them had previously performed at least 1000 procedures. The study population consisted of serially recruited patients 18 years scheduled to undergo ERCP and, when indicated, endoscopic sphincterotomy (Table 1). Exclusion criteria were active or chronic pancreatitis; cancers of pancreatic, ampullary, or biliary origin; and a previous sphincterotomy. All eligible patients were invited to take part in the study and had to give written informed consent before study entry. The study was approved by ethics committees at each center. Study Design Data were collected by means of a standardized questionnaire at the time of the procedure, before discharge, and 30 days afterward. Patients who met criteria for entry were randomized in blocks of 30 patients at each center according to a computer-generated list. All ERCPs were performed on hospital-admitted patients. Preparing and starting an infusion of either placebo or experimental drugs were performed in the Endoscopic Unit, whereas physicians caring for patients in the ward were unaware of assigned treatments. By this policy, blindness about treatment options was ensured. Placebo (500

3 August 2004 PROPHYLAXIS OF ERCP-RELATED PANCREATITIS 715 ml of 0.9% saline solution), somatostatin (750 g dissolved in 500 ml of saline solution), and gabexate (500 mg in 500 ml of 5% glucose solution) were infused intravenously, starting 30 minutes before the endoscopic session and continuing for 6 hours afterward. Therapy with antibiotics, analgesics, and sedatives was allowed. At the end of each procedure, the endoscopist recorded procedural details of the maneuvers performed, particularly filling of the biliary and/or pancreatic ducts, and their morphological characteristics (whether strictured or normal, with delayed or normal emptying); number of cannulations; number of pancreatic duct injections; length of sphincterotomy, measured taking the 2-cm length of a standard sphincterotome as internal reference; whether a needleknife sphincterotomy was performed; diameter of the bile duct; and the presence of choledocolithiasis. Finally, the level of technical difficulty inherent to the procedure was graded according to the scores of both Freeman et al. 9 and Schutz and Abbott. 10 Immediately after endoscopy, patients returned to the hospital ward and fasted for at least 8 hours. Serum amylase concentrations were measured before endoscopy and at 4, 24, and 48 hours afterward. The presence of abdominal pain and use and type of analgesics also were recorded at the same times. Definitions Patients were considered to have post-ercp pancreatitis when both abdominal pain requiring analgesic use and lasting for at least 24 hours after endoscopy and a persistent increase in serum amylase levels occurred. Following established criteria, 11,12 serum amylase level was considered significantly increased when 5- and 3-fold elevations above the upper limits of normal at 4 and 24 hours after the procedure were recorded, respectively. All patients with post-ercp pancreatitis underwent abdominal ultrasound to search for peripancreatic fluid collection. Computed tomography was performed only in case of a failed pancreatic ultrasound examination. Severity of post-ercp pancreatitis was graded according to ultrasound or computed tomographic findings: the presence of either tissue necrosis involving 30% of the pancreatic gland or peripancreatic fluid collection classified pancreatitis as severe. 13 Asymptomatic hyperamylasemia is defined as a 2-fold elevation in serum amylase levels above the upper limits of normal, associated or not with abdominal pain that did not require analgesics. On the basis of serum amylase levels at the 24-hour evaluation after endoscopy, these cases were sorted further into transient (normal or 2-fold increase) or persistent ( 2-fold increase) hyperamylasemia. A postendoscopic decrease in serum hemoglobin level 2 g/dl from baseline was considered evidence of bleeding, whereas new onset of fever accompanied by an increase in white blood cell count and liver biochemical test results was diagnosed as cholangitis. Statistical Analysis A sample size of 1050 endoscopic procedures (350 in each experimental group) was estimated to have to be performed in the trial on an type error of 0.05 and error of 0.20 and on the assumption of a 6% 8% incidence of pancreatitis in controls and 3% 5% incidence in patients treated with proactive drugs. The primary end point was a reduction in the incidence of pancreatitis after ERCP. To improve consistency of data collection within this multicenter study, data were blindly validated by an Independent Data Monitoring Committee at the coordinating center. Committee members (A.A., F.P, V.F.) were not personally involved in performing the endoscopic procedures or in charge of the included patients. Chi-square and Fisher s exact tests for comparison of categorical data were used, and P 0.05 is considered statistically significant. Potentially relevant risk factors were assessed by means of univariate analysis, and significant predictors were included in logistic regression analysis to identify the most important risk factors for the considered end point. Results A total of 1127 endoscopic procedures were performed: 351 with the prophylactic administration of somatostatin; 381, with gabexate; and the remaining 395, with placebo. Treatment groups were similar with regard to sex, mean age, indications for endoscopy, occurrence of risk factors, and difficulty of ERCP (Table 1). Incidence of Post-ERCP Pancreatitis Acute pancreatitis developed in 63 patients after ERCP, with an overall incidence of 5.6% (Table 2). Patients treated with somatostatin and gabexate had a greater incidence of pancreatitis than those treated with placebo (6.3% and 5.8% vs. 4.8%, respectively), but the difference was not significant (Table 2). Acute pancreatitis was clinically mild and edematous, as assessed by ultrasonography or computed tomography, in 58 patients (92.0% of 63 cases of pancreatitis; 5.1% of all procedures) and severe in 4 cases, all of them treated with gabexate. Results of the analysis of risk factors for pancreatitis are listed in Table 3. Of 21 variables evaluated, 2 variables were significant at both univariate and multivariate analysis and were related to the difficulty in obtaining biliary access: 3 contrast medium injections into the pancreas and a Freeman score 1. Of the remaining variables, only precut sphincterotomy was significant at the univariate analysis. Incidence of Hyperamylasemia and Abdominal Pain Before ERCP, 22 patients had a serum amylase level 5-fold the upper limits of normal. Excluding patients with pancreatitis, hyperamylasemia ( 2-fold the upper limits of normal) and abdominal pain were

4 716 ANDRIULLI ET AL. CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 8 Table 2. Incidence of Post Endoscopic Retrograde Cholangiopancreatography Events in 1127 Patients, Subdivided According to Experimental Treatment Placebo Gabexate Somatostatin Total No. of patients No pain and normal amylase 226 (57.3) 215 (56.4) 217 (61.8) 658 (58.3) Abdominal pain with normal amylase 21 (5.3) 24 (6.3) 18 (5.1) 63 (5.6) Transient hyperamylasemia 53 (13.4) 33 (8.7) 32 (9.1) 118 (10.5) Persistent hyperamylasemia 76 (19.2) 87 (22.8) 62 (17.7) 225 (20.0) Acute pancreatitis 19 (4.8) 22 (5.8) 22 (6.3) 63 (5.6) Cholangitis 2 (0.5) 5 (1.3) 3 (0.9) 10 (0.9) Bleeding 7 (1.8) 7 (1.8) 5 (1.1) 19 (1.7) Perforation 2 (0.5) 1 (0.3) 1 (0.3) 4 (0.4) Death 4 (1.0) 3 (0.8) 0 7 (0.6) observed after 343 (30.5%) and 63 procedures (5.6%), respectively (Table 2). Persistent hyperamylasemia at 24 hours after the procedure was found after 225 ERCPs (20.0%), whereas transient hyperamylasemia, present at 4 hours, but not at 24 or 48 hours, was found in 118 patients (10.5%). Rates of either persistent or transient hyperamylasemia did not differ among the 3 treatment groups (Table 2). Other Complications Twenty-three post-ercp complications, in addition to pancreatitis, were observed in 16 patients, at similar rates among the 3 treatment groups. Cholangitis was found in 10 patients; bleeding, in 18 patients; retroduodenal perforations, in 3 patients; and esophageal perforation, in 1 patient. Seven patients died within 30 days after ERCP; 4 and 3 deaths were registered after placebo and gabexate, respectively. Five deaths were directly related to the endoscopic procedure because they followed severe gastrointestinal bleeding (1 patient), retroduodenal perforation (1 patient), septic shock secondary to cholangitis (1 patient), and acute pancreatitis of the necro-inflammatory hemorrhagic (1 patient) or mild type (1 patient). The remaining 2 patients with severe systemic disease (scores 4 and 5 of the American Society of Anesthesiology classification) died of pulmo- Table 3. Risk Factors for Pancreatitis After ERCP at Univariate and Multivariate Analyses Risk factors Patients with pancreatitis (n 63) Patients without pancreatitis (n 1064) P Odds ratio (95% CI) Significant at both analyses Freeman score Grade 1 24 (38.1) 689 (64.8) Grade (61.9) 375 (35.2) 1.47 ( ) Pancreatic contrast injections 0, 1, or 2 41 (65.1) 962 (90.4) (34.9) 102 (9.6) 1.95 ( ) Significant at univariate analysis only Precut sphincterotomy 10 (15.9) 79 (7.4) ( ) Not significant Female sex 36 (57.1) 513 (48.2) 0.17 Age 35 yr 6 (8.0) 69 (6.5) 0.34 Choledocolithiasis 46 (73.0) 820 (77.1) 0.46 Stenosis of the main bile duct 15 (23.8) 277 (26.0) 0.69 Cholangitis 5 (7.9) 156 (14.7) 0.14 Biliary pain 36 (57.1) 566 (53.2) 0.54 History of acute pancreatitis 10 (15.9) 183 (17.2) 0.79 History post-ercp pancreatitis 1 (1.6) 12 (1.1) 0.74 Common bile duct diameter 8 mm 38 (60.3) 708 (66.5) 0.31 Suspected sphincter of Oddi dysfunction 4 (6.3) 59 (5.5) 0.99 Impacted or stenotic papilla 10 (15.9) 103 (9.7) 0.25 Sphincterotomy 47 (74.6) 810 (76.2) 0.78 Length of sphincterotomy 2 cm a 47 (100) 767 (94.7) 0.11 Acinarization of pancreas 3 (4.8) 22 (2.1) 0.17 NOTE. Only risk factors with P 0.05 at univariate analysis were included in multivariate analysis. ERCP, endoscopic retrograde cholangiopancreatography; CI, confidence interval. a Percentage values refer to 857 patients who underwent successful sphincterotomy.

5 August 2004 PROPHYLAXIS OF ERCP-RELATED PANCREATITIS 717 nary embolism (1 patient) and cardiopulmonary failure (1 patient), causes thought to be unrelated to the procedure itself. Discussion In the current trial, we provide evidence that an intravenous infusion of either somatostatin or gabexate, starting 30 minutes before ERCP and lasting for 6 hours thereafter, was ineffective in reducing the incidence of post-ercp hyperamylasemia, abdominal pain, and pancreatitis. The present study is in agreement with similar negative results observed after a 2.5-hour infusion of the 2 drugs in high-risk patients undergoing ERCP and reported in our previous collaborative study (the GIPPA 1 trial). 5 Moreover, the 2 experimental drugs also were ineffective in reducing other complications, such as cholangitis and bleeding, as well as the number of patients who died after the endoscopic procedure. The main argument to dispute results of the GIPPA 1 trial is the short time of drug application: the 2.5-hour infusion time might have been able to block secretion of enzymes stored in acinar cells, but not prevent pancreatitis, which has been reported to start with a delay of 4.5 hours after the procedure. 8 This claim is disproved by data from the present GIPPA 2 study on clinical grounds, as well as by experimental evidence. In a human model of post-ercp pancreatitis, it was shown that the cytokine-induced acute phase of response to the pancreatic injury starts immediately after the damage, well before the onset of clinical symptoms of pancreatitis. 8,14 16 The implication of these data is that the pancreatic damage is set very early in the course of the endoscopic procedure, and attempts to modify this series of events must be undertaken before or very soon after the noxious insult occurs. Several recent studies reported that the incidence of post-ercp pancreatitis could be reduced even after a single intravenous prophylactic dose of interleukin and somatostatin 18 or a single suppository of diclofenac. 19 After disproving the argument that a difference in infusion times might explain the divergent results between our data and those of other prophylactic trials, we acknowledge 2 recent trials reporting opposite results from our present data. In the first trial, somatostatin was administered as a bolus, and the occurrence of clinical pancreatitis (4.4% vs. 13.3%) and hyperamylasemia (26.0% vs. 38.5%) were both significantly lower than after placebo. 18 In the second study, a short infusion of gabexate (6.5 hr) was equally protective as a longer infusion (12 hr). 6 How can we reconcile these differences? By scrutinizing the trial designs, no major difference in either patient characteristics or procedural maneuvers was apparent from both the GIPPA 1 and 2 studies and other trials. We may only observe a disproportionately high rate of pancreatitis in the control groups from studies reporting positive data: 22% in the interleukin trial, 17 18% in the methylprednisolone trial, % in the diclofenac study, % in the study with bolus somatostatin, 18 and 8% in the gabexate trial. 2 Conversely, most trials with negative results, 4,21 including the present trial, reported a pancreatitis rate 7% in the placebo-treated groups, in accordance with the 1% 7% expected rate of ERCP-induced pancreatitis quoted by the Standards of Practice Committee of the American Society for Gastrointestinal Endoscopy. 22 Because the distinction between clinical mild pancreatitis and transient abdominal pain with a marginal increase in serum amylase levels is arbitrary, the recommendation has been issued to randomly validate experimental data by second investigators not involved in the original collection of data. 23 With the intent to avoid either overdiagnosis or underdiagnosis of post-ercp complications, an independent data monitoring committee validated data in both the present and our previous study. 5 Unfortunately, a similar committee was not consistently in place in all previous studies. Numerous patient- and technique-related factors are commonly quoted as related to the development of pancreatitis. However, in our GIPPA 1 study, conditions linked to the occurrence of post-ercp hyperamylasemia were number of pancreatic contrast injections and sphincterotomy; those significantly related to pancreatitis were sphincterotomy, unsuccessful bile duct cannulation, and number of pancreatic contrast injections. The present study re-emphasizes that much of the injury to the pancreas appears to result from the process of cannulating the bile duct, rather than from baseline characteristics of patients: at multiple logistic regression analysis, 3 pancreatic injections (OR, 1.95; 95% CI, ) and a Freeman score 1 (OR, 1.47; 95% CI, ) were predictive of post-ercp pancreatitis. Therefore, because most endoscopists participating in these trials were highly experienced, current predictive factors for post-ercp pancreatitis are different from those traditionally mentioned and are related mainly to the number of pancreatic duct opacification. In conclusion, in the current study, patients were administered a 6.5-hour infusion of either somatostatin or gabexate before ERCP, which should allow for longlasting preventive effects. The negative results of both the GIPPA 1 and 2 trials are evidence for not recommending the 2 drugs for the prophylaxis of pancreatic

6 718 ANDRIULLI ET AL. CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 8 damage after ERCP. Future prophylactic interventions might be needed only for patients with several pancreatic duct opacifications during the endoscopic procedure. References 1. Andriulli A, Caruso N, Quitadamo M, Forlano R, Leandro G, Spirito F, De Maio G. Antisecretory vs. antiproteasic drugs in the prevention of post-ercp pancreatitis: the evidence-based medicine derived from a meta-analysis study. J Pancreas 2003;4: Cavallini G, Tittobello A, Frulloni L, Masci E, Mariani A, Di Francesco V, and the Gabexate-Mesilate in Digestive Endoscopy Italian Group. Gabexate for the prevention of pancreatic damage related to endoscopic retrograde cholangiopancreatography. N Engl J Med 1996;335: Poon RTP, Yeung C, Lo CM, Yuen WK, Lin CL, Fan ST. Prophylactic effect of somatostatin on post-ercp pancreatitis: a randomized controlled trial. Gastrointest Endosc 1999;49: Andriulli A, Leandro G, Niro G, Mangia A, Festa V, Gambassi G, Villani MR, Facciorusso D, Conoscitore P, Spirito F, De Maio G. Medical treatment can prevent pancreatic injury after ERCP: a meta-analysis. Gastrointest Endosc 2000;51: Andriulli A, Clemente R, Solmi L, Terruzzi V, Suriani R, Sigillito A, Leandro G, Leo P, De Maio G, Perri F. Gabexate or somatostatin administration before ERCP in patients at high risk for post-ercp pancreatitis: a multicenter, placebo-controlled, randomized clinical trial. Gastrointest Endosc 2002;56: Masci E, Cavallini G, Mariani A, Frullini L, Testoni PA, Curioni S, Tittobello A, Uomo G, Costamagna G, Zimbelli S, Macarri G, Innocenti P, Dragonetti C. Comparison of two dosing regimens of gabexate in the prophylaxis of post-ercp pancreatitis. Am J Gastroenterol 2003;98: Masci E, Mariani A. Short-term prophylactic prevention of post- ERCP pancreatitis. Gastrointest Endosc 2003;58: Messmann H, Vogt W, Holstege A, Lock G, Heinisch A, von Furstenberg A, Leser HG, Zirngible H, Scholmerich J. Post-ERCP pancreatitis as a model for cytokine induced acute phase response in acute pancreatitis. Gut 1997;40: Freeman ML, Nelson DB, Sherman S, Haber GB, Herman ME, Dorsher PJ, Moore JP, Fennerty MB, Ryan ME, Shaw MJ, Lande JD, Pheley AM. Complications of endoscopic biliary sphincterotomy. N Engl J Med 1996;335: Schutz SM, Abbott RM. Grading ERCPs by degree of difficulty: a new concept to produce more meaningful outcome data. Gastrointest Endosc 2000;51: Cotton PB, Lehman G, Vennes J, Geenen JE, Russell RC, Meyers WC, Liguory C, Nickl N. Endoscopic sphincterotomy complications and their management: an attempt at consensus. Gastrointest Endosc 1991;37: Testoni PA, Bagnolo F. Pain at 24 hours associated with amylase levels greater than 5 times the upper normal limit as the most reliable indicator of post-ercp pancreatitis. Gastrointest Endosc 2001;53: Balthazar EJ, Robinson DL, Megibow AJ, Ranson JH. Acute pancreatitis: value of contrast tomography in establishing prognosis. Radiology 1990;174: Geenen JE, Hogan WJ, Dodds WJ. Sphincter of Oddi. In: Sivak MV, ed. Gastroenterologic Endoscopy. Philadelphia, PA: Saunders, 1987: Leser HG, Gross V, Scheibenbogen C, Heinisch A, Salm R, Lausen M, Ruckauer K, Andreesen R, Farthmann EH, Scholmerick J. Elevation of interleukin-6 concentration precedes acutephase response and reflects severity in acute pancreatitis. Gastroenterology 1991;101: Saluja AK, Steer ML. Pathophysiology of pancreatitis. Digestion 1999;60(suppl 1):S27 S Deviere J, Le Moine O, Van Laethem JL, Eisendrath P, Ghilain A, Severs N, Cohard M. Interleukin-10 reduces the incidence of pancreatitis after therapeutic endoscopic retrograde cholangiopancreatography. Gastroenterology 2001;120: Poon RTP, Yeung C, Liu CL, Lam CM, Yen WK, Lo CM, Fan ST. Intravenous bolus somatostatin after diagnostic cholangiopancreatography reduces the incidence of pancreatitis associated with therapeutic endoscopic retrograde cholangiopancreatography procedures: a randomised controlled trial. Gut 2003;52: Murray B, Carter R, Imrie C, Evans S, O Suilleabhain C. Diclofenac reduces the incidence of acute pancreatitis after ERCP. Gastroenterology 2002;124: Dumot JA, Conwell DL, O Connor JB, Ferguson R, Vargo JJ, Barnes DS, Shay SS, Sterling MJ, Issa K, Ponskky JL, Zuccaro G. Pretreatment with methylprednisolone to prevent ERCP-induced pancreatitis: a randomized, multicentre, placebo-controlled clinical trial. Am J Gastroenterol 1998;93: Dumot JA, Darwin CL, Zuccaro G, Vargo JJ, Shay SS, Kirk A, Easley MS, Ponskky JL. A randomized, double-blind study of interleukin 10 for the prevention of ERCP-induced pancreatitis. Am J Gastroenterol 2001;96: Standards of Practice Committee of the American Society for Gastrointestinal Endoscopy. Complications of ERCP. Gastrointest Endosc 2003;57: Freeman M. Toward improving outcomes of ERCP. Gastrointest Endosc 1998;48: Address requests for reprints to: Angelo Andriulli, M.D., Division of Gastroenterology, Casa Sollievo della Sofferenza Hospital, IRCCS, San Giovanni Rotondo, Italy. a.andriulli@operapadrepio.it; fax: (39) The trial was designed and the manuscript was written by Dr. Andriulli. No pharmaceutical or private grants were searched for or made available for performing the present work.