endoscopy corner - Clinical Gastroenterology and Hepatology

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2011;9:851– 858

ENDOSCOPY CORNER Endoscopic Pancreatic Duct Stents Reduce the Incidence of Post–Endoscopic Retrograde Cholangiopancreatography Pancreatitis in High-Risk Patients ATSUSHI SOFUNI,* HIROYUKI MAGUCHI,‡ TSUYOSHI MUKAI,§ HIROSHI KAWAKAMI,储 ATSUSHI IRISAWA,¶ KENSUKE KUBOTA,# SHINJI OKANIWA,** MASATAKA KIKUYAMA,‡‡ HIROMU KUTSUMI,§§ KEIJI HANADA,储储 TOSHIHARU UEKI,¶¶ and TAKAO ITOI* *Department of Gastroenterology and Hepatology, Tokyo Medical University, Tokyo; ‡Center for Gastroenterology, Teine-Keijinkai Hospital, Hokkaido; §Department of Gastroenterology, Gifu Municipal Hospital, Gifu; 储Department of Gastroenterology, Hokkaido University Graduate School of Medicine, Hokkaido; ¶Preparatory office for Aizu Medical Center, Fukushima Medical University, Fukushima; #Department of Gastroenterology, Yokohama City University, Kanagawa; **Department of Gastroenterology, Iida Municipal Hospital, Nagano; ‡‡Department of Gastroenterology, Hamamatsu Rosai Hospital, Shizuoka; §§Department of Gastroenterology, Kobe University, Kobe; 储储Center for Gastroendoscopy, Onomichi General Hospital, Hiroshima; and ¶¶Department of Gastroenterology, Fukuoka University Chikushi Hospital, Fukuoka, Japan

This article has an accompanying continuing medical education activity on page e110. Learning Objectives—At the end of this activity, the learner should recognize the utility of pancreatic ductal stenting in reducing the rate of post-ERCP pancreatitis, appreciate the risk factors associated with the development of post-ERCP pancreatitis, understand that these risk factors may be additive further emphasizing the importance of pancreatic ductal stenting, and learn the technical aspects of the procedure, including the type of stents that are used.

Scan the quick response (QR) code to the left with your mobile device to watch this article’s video abstract and others. Don’t have a QR code reader? Get one at mobiletag.com/en/download.php.

See editorial on page 810. BACKGROUND & AIMS: Pancreatitis is the most common and potentially serious complication of post-endoscopic retrograde cholangiopancreatography (ERCP). Post-ERCP pancreatitis (PEP) is caused mostly by postprocedural papillary edema and retention of pancreatic juice. We conducted a randomized controlled trial to determine whether placement of a temporary-type, pancreatic duct stent prevents PEP and to identify risk factors for PEP. METHODS: We analyzed data from 426 consecutive patients who underwent ERCP-related procedures at 37 endoscopic units. The patients were assigned randomly to groups that received stents (S group, n ⫽ 213) or did not (nS group, n ⫽ 213). The stent used was temporary, 5F in diameter, 3 cm long, and straight with an unflanged inner end. RESULTS: The overall frequency of PEP was 11.3%. The frequencies of PEP in the S and nS groups were 7.9% and 15.2%, respectively; the lower incidence of PEP in the S group was statistically significant based on the full analysis set (P ⫽ .021), although there was no statistically significant differences in an intention-to-treat analysis (P ⫽ .076). There were significant differences in PEP incidence between groups in multivariate analysis for the following risk factors: pancreatography first, nonplacement of a pancreatic duct stent after ERCP, procedure time of 30 minutes or more, sampling of pancreatic tissue by any method, intraductal ultrasonography, and difficulty of cannulation (ⱖ15 min). Patients with more than 3 risk factors had a significantly greater incidence of pancreatitis. CONCLUSIONS: Placement of a pancreatic duct stent reduces the incidence of PEP. Several risk factors are associated with PEP. Keywords: Inflammation; Pancreas; Clinical Trial; Drainage.

Watch this article’s video abstract and others at http:// tiny.cc/bz9jv.

P

ost–endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis is recognized as the most common and potentially serious complication of ERCP. The frequency of postERCP pancreatitis (PEP) is reported to be from 1% to 9%, and more than 11% to 40% in cases with risk factors, according to recent large prospective trials.1–26 Although various mechanisms of post-ERCP pancreatitis have been suggested,1–26 one of the most frequent mechanisms is impaired drainage from the pancreatic duct caused by papillary edema and/or spasm of the sphincter of Oddi after the procedure.10 –14,19,22,25,26 As a countermeasure, several endoscopists have inserted a nasopancreatic drainage tube into the pancreatic duct9,10 or used a flanged pancreatic stent.10 –14,19,20 A pancreatic spontaneous dislodgement duct stent (PSDS) recently has become commercially Abbreviations used in this paper: C-IDUS, common bile duct– intraductal ultrasonography; ERCP, endoscopic retrograde cholangiopancreatography; ES, endoscopic sphincterotomy; FAS, full analysis set; IDUS, intraductal ultrasonography; ITT, intention-to-treat; PEP, post– endoscopic retrograde cholangiopancreatography pancreatitis; P-GW, pancreatic guidewire placement; P-IDUS, pancreatic duct intraductal ultrasonography; PS, pancreatic duct stent; PSDS, pancreatic spontaneous dislodgement duct stent; RCT, randomized controlled multicenter trial. © 2011 by the AGA Institute 1542-3565/$36.00 doi:10.1016/j.cgh.2011.06.033

852

SOFUNI ET AL

available in Japan. Pancreatic duct stents (PS) are classified into those with or without flanges on the pancreatic ductal side. The former type is less likely to dislodge naturally, and endoscopic removal is required more often. Regarding the unflanged PS, the rate of natural dislodgement within a short period is high,1,17,25,26 and re-insertion of an endoscope for removal generally is unnecessary. Recently, PSs have been reported to be effective in preventing PEP.1,15,17–20,25–30 Moreover, we reported the effectiveness of PSDS in preventing PEP based on the results of a randomized controlled multicenter trial (RCT) in 2006.26 This trial included cases that met inclusion criteria and needed ERCP, to evaluate whether PSDS was effective. It clearly showed the efficacy of PSDS, but it was also important in reviewing risk factors according to the kind of case in which PSDS was used. We conducted an RCT to evaluate whether placement of a PSDS would prevent pancreatitis after ERCP-related procedures, in patients with any risk factors of PEP, and to identify risk factors of PEP.

Materials and Methods Materials This RCT was performed between April and June 2008 at 37 large endoscopic units in Japan, each performing more than 300 ERCPs per year. The sample size calculations indicated that a total of approximately 268 patients would be required to detect a decrease in the incidence of PEP from 15% to 5% (␣ ⫽ .01, ␤ ⫽ .9; 2-tailed test).1–17 We enrolled 426 patients to allow for exclusion, invalid, and protocol deviation cases from 37 institutions to have sufficient statistical power to compare the severity of PEP and risk factors with our previous RCT results. A total of 426 consecutive patients underwent ERCP-related procedures. The patients were randomized in advance by randomization software into a stent placement group (S group, 213 patients), or a nonstent placement group (nS group, 213 patients), and then notified using opaque, sealed envelopes distributed by a third person. The patients were evaluated by observers who were blinded as to whether they had a stent or not (Supplementary Appendix 1). The study group included patients who had risk factors of PEP. The selection of risk factors, and the setting of inclusion and exclusion criteria, were determined after the consideration of various previous studies1–30 and discussions by the collaborating organizations (Supplementary Appendix 2). Risk factors were divided into 2 groups: (1) patient-related factors and (2) procedure-related factors. Patient-related factors included the following: (1) age younger than 60 years and female sex, (2) history of pancreatitis, and (3) suspected biliary sphincter of Oddi dysfunction. Procedure-related factors included the following: (1) pancreatography first, (2) pancreatography 2 or more times, (3) endoscopic pancreatic sphincterotomy, (4) precut sphincterotomy, (5) endoscopic papillary balloon dilation, (6) common bile duct (C) tissue sampling (biopsy, brush, cytology), (7) pancreatic duct (P) tissue sampling (biopsy, brush, cytology), (8) endoscopic biliary drainage without endoscopic sphincterotomy (ES), (9) endoscopic nasobiliary drainage without ES, (10) common bile duct–intraductal ultrasonography (C-IDUS), (11) pancreatic duct–IDUS (P-IDUS), (12) difficulty of cannulation (ⱖ15 min), and (13) procedure time (ⱖ30 min).

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 10

The following cases were excluded: (1) cases who could not provide written informed consent, (2) cases deemed inappropriate for this trial, (3) cases with a performance status of 4 (completely disabled, could not perform any self-care, totally confined to bed or chair), (4) cases in which the duodenal papilla could not be accessed endoscopically, (5) cases requiring pancreatic duct drainage, (6) cases with a history of prior ERCP procedures, (7) cases of postgastrectomy, (8) cases requiring endoscopic papillectomy, (9) cases of pancreatic head cancer with main pancreatic duct occlusion, (10) cases of pancreas divisum (excluding such cases at the stage when diagnosed by ERCP because of the difficulty of diagnosis before pre-ERCP), (11) cases of chronic pancreatitis, (12) cases of intraductal papillary mucinous neoplasms, (13) cases of gallstone pancreatitis, (14) cases of pancreaticobiliary maljunction, and (15) cases in which the ERCP procedure was not completed within 1 hour from the transoral insertion of the scope.

Study Design Before endoscopy, the history of the patient was taken and a physical examination was performed. Pharyngeal anesthesia was induced with a topical anesthetic and conscious sedation by intravenous medication. We gave a proteolytic enzyme, ulinastatin (50,000 U), dissolved in a 500-mL solution and administered by continuous intravenous infusion immediately after ERCP procedures and thereafter every 12 hours, according to the guidelines for initial treatment based on the acute pancreatitis guide to medical care by the Ministry of Health, Labor and Welfare, Intractable Pancreatic Disease Research Group, Japan. Antibiotics also simultaneously were administered by drip infusion. The kinds of ERCP devices used (ie, catheter, sphincterotome, or guidewires), were not limited to any specific types. For stent placement, we mostly used the Radifocus (Terumo, Co, Ltd, Tokyo, Japan) guidewire or the Jagwire (Boston Scientific Japan, Tokyo, Japan) guidewire. In this study, contrast agent was injected first in the ERCP procedures, as is the custom in Japan. Sphincterotomy was performed with a standard traction sphincterotome and alternating electrosurgical cutting and coagulation current. In cases in which it was difficult to cannulate the bile duct in ERCP procedures, we performed pancreatic guidewire placement (P-GW) to achieve selective biliary cannulation.27,28 We used a 5F straight polyethylene stent, 3 cm in length, which was unflanged on the pancreatic ductal side but with 2 flanges on the duodenal side (GPDS-5-3; Cook Endoscopy, Inc, Winston-Salem, NC). Stent dislodgment was confirmed by daily abdominal radiographs, taken every morning until confirmation of dislodgement. The stent was planned to be removed by duodenoscopy on the morning of the fourth day, if it had not dislodged by the third day. The randomization of patients was performed on enrollment. The group assigned to stent placement (S group), first underwent ERCP and related procedures (biliary or pancreatic procedures). After those procedures, the placement of PSDS was performed. The stent was passed over a guidewire under fluoroscopic guidance. In the nS group, the procedure was finished without placement of a stent. When a case was randomized into the nS group, the procedure was finished without cannulation or injection of the pancreatic duct, after the biliary procedures, if it was a routine biliary case. After the procedures,

October 2011

PANCREATIC DUCT STENTS AND PEP

853

Table 1. Summary of Stent Placement ITT (n ⫽ 213) Rate of success in stent placement (%) Rate of spontaneous stent dislodgement (%) Rate of removal endoscopically (%) Mean duration time to dislodgement, d (range) Adverse events Stent migration (%) Post-ERCP pancreatitis (%) Hyperamylasemia (%) Mean serum amylase levels after procedures, U/L (range) Hemorrhage (%) Perforation (%) Infection (cholangitis, cholecystitis) (%) Others (%) 25 unsucessful stent placement cases, mild pancreatitis (%)

the endoscopists recorded the results. After ERCP, patients fasted until blood tests confirmed there was no pancreatitis or other complications on the following day. All patients were hospitalized for the ERCP procedure and observation. All patients provided written informed consent before entry into the study. Institutional Review Boards at each institution reviewed and approved the study.

Definitions The definition of PEP was standardized by a consensus conference held in 1991, and those criteria have become widely accepted.1,22 PEP was defined as pancreatic pain and hyperamylasemia within 24 hours of the procedure. Pancreatic pain was defined as persistent pain in the epigastric or periumbilical region. Hyperamylasemia was defined as an increase in serum amylase to more than 3 times the upper normal limit defined by each institution. We assessed the analysis of the amylase and pain in every 24-hour period after ERCP. The definition of severity was modified based on the criteria of Cotton et al.22 However, although the Cotton et al22 criteria defined the degree of severity as days of hospitalization,2–10 each institution used the same number of days of fasting.

Outcomes The primary study end points were the frequency and severity of PEP in patients with risk factors. We also evaluated hyperamylasemia, the success rate of stent placement, the duration before stent dislodgment, and complications in both groups. We also evaluated risk factors for PEP. Each endoscopist had performed more than 300 ERCP procedures per year for 2 to 20 years. It also was recorded if the endoscopist changed during the session.

Statistical Analysis Statistical analysis was performed on the basis of intention-to-treat (ITT) analysis. Drop-out cases should not be excluded from the applicable analytic cases in our RCT. An analysis of all cases except for excluded cases comprises the full analysis set (FAS), which reflects the clinical situations. The results of the FAS also were shown together with those of ITT analysis in this study.

188/213 (88.3) 178/188 (95.2) 10 (4.7) 1.8 (0–4) 0 (0) 20 (9.4) 126 (60) 336.3 (1.7–2767) 0 (0) 0 (0) 0 (0) 0 (0) 3/25 (12.0)

FAS (n ⫽ 203) 186/203 (91.6) 178/186 (95.7) 10 (4.9) 1.8 (0–4) 0 (0) 16 (7.8) 124 (60) 328.0 (1.7–2767) 0 (0) 0 (0) 0 (0) 0 (0) 3/23 (13.0)

The ␹2 test (with Yates correction) or the Fisher exact test was used to determine the significance of associations. A P value of less than .05 indicated a statistically significant difference. Univariate evaluation was first made for each potential risk factor using the chi-square test. In the second step, only the significant factors (P ⬍ .10) on chi-square analysis were included on multivariate (logistic regression) analysis. The Cochran–Armitage test was used as a trend test.

Results As for the final details of all enrolled 426 cases (S group, 213 cases; nS group, 213 cases), 12 cases did not meet inclusion criteria and 7 other cases had incomplete data sheets, thus 407 cases were able to be analyzed (S group, 203 cases; nS group, 204 cases). The competence cases excluded 12 cases of the following: protocol deviation case (ie, a pancreatic cancer, chronic pancreatitis, and pancreatic divisum case) to conform to exclusion criteria, cases deemed inappropriate for this trial such as cases of difficulty in judgment of abdominal pain, and cases in which the PSDS placement did not follow the protocol. Unsuccessful cases of stent placement were not excluded from both analyses because the failure of stent placement can happen in the clinical setting (Supplementary Appendix 1). Statistically, FAS reflected actual clinical situations and has clinical significance. Therefore, the results of the FAS also were shown together with those of ITT analysis in this study. Table 1 shows the final status of patients with risk factors in both groups in whom ERCP was indicated. Indications and other parameters were generally no different in the 2 groups (Supplementary Appendix 3).

Pancreatic Duct Stent Placement The success rate of PS placement was 88.3% on ITT analysis (91.6% on FAS). The rate of spontaneous dislodgment by day 3 was 95.2% (95.7% on FAS) and the stent had to be removed endoscopically in 4.7% (4.9% on FAS) cases. The mean duration to dislodgment was 1.8 days (range, 0 – 4 d) (Table 1). No major complications such as stent migration, hemorrhage, perforation, or biliary severe infection as a result of stent placement were seen. The only complications in 25 unsuccessful PS

854

SOFUNI ET AL


Table 2. PEP Rates, Severity, and Mean Amylase Levels S group Patients, n PEP (%) Severity of PEP Mild Moderate Severe Mean serum amylase level after procedures in pancreatitis case Hyperamylasemia Mean serum amylase level, IU/L Abdominal pain Overall mean serum amylase levels after procedures

Patients, n PEP (%) Severity of PEP Mild Moderate Severe Mean serum amylase level after procedures in pancreatitis case Hyperamylasemia Mean serum amylase level, IU/L Abdominal pain Overall mean serum amylase levels after procedures

placement cases were mild pancreatitis, and the rate of frequency was 12.0% (13.0%). The cause of failures in this study were as follows: there were 3 cases of anatomic deformities of the pancreatic duct, 15 cases of inability to cannulate the pancreatic duct, and 7 cases of difficulty passing the guidewire deeply enough.

Post–Endoscopic Retrograde Cholangiopancreatography Pancreatitis The mean registration number of cases per 1 institution was 11.5 (range, 2–35). The mean frequency of PEP in each institution was 11.2 (0%– 60%), but there was a significant difference in the frequency of PEP among the facilities (P ⫽ .02). The overall frequency of PEP was 11.9% (51 of 426) on ITT analysis, and 11.3% (46 of 407) on FAS. The frequency of PEP in the S and nS groups was 9.4% (20 of 213) and 14.6% (31 of 213), respectively, showing a lower frequency of PEP in the S group (P ⫽ .076, ITT analysis). On FAS, the frequency of PEP in the S and nS groups was 7.9% (16 of 203) and 15.2% (31 of 204), respectively, showing a statistically significant lower frequency of PEP in the S group (P ⫽ .021). Table 2 shows the frequency and severity of PEP in the S and nS groups. There was no statistically significant difference in the frequency of severe pancreatitis between the S and nS groups because of the limited number of pancreatitis cases. The mean increase in amylase level in the pancreatitis patients also was not significant between the S and nS groups. There was no statistically significant difference in the frequency of hyperamylasemia, mean serum amylase of hyperamylasemia, abdominal pain, or in the overall mean serum amylase after procedures between the S and nS groups.

213 20 (9.4)

nS group

P value (ITT)

213 31 (14.6)

.076

16 4 0 1257 (453–2469) 88 (41.9%) 388.1 (142–2767) 37 (17.6%) 336.3 (1.7–2767)

22 8 1 1400 (312–2933) 70 (33.0%) 312.2 (143–967) 38 (17.9%) 349.3 (16–2933)

.240 .389 1.000 .491 .996 .059 .935 .885

S group

nS group

P value (FAS)

203 16 (7.9) 12 4 0 1300 (453–2469) 86 (42.4%) 376.8 (142–2767) 33 (16.3%) 328.0 (1.7–2767)

204 31 (15.2) 22 8 1 1400 (312–2933) 68 (33.3%) 180.5 (143–967) 38 (17.9%) 358.5 (16–2933)

.021 .770 .952 1.000 .797 .200 .070 .528 .554

Risk Factors for Post–Endoscopic Retrograde Cholangiopancreatography Pancreatitis The incidence rate of PEP and the results of univariate analysis of patient-related and procedure-related risk factors were analyzed (Supplementary Appendix 4). There was no difference in the number of pancreatography procedures per patients (ⱖ2). However, there was a statistically significant difference between the mean number of pancreatography procedures in the pancreatitis group and no pancreatitis group (4.84 vs 2.84 times; P ⫽ .04). Table 3 shows the significant risk factors for PEP on univariate analysis, that is, pancreatography first, P-tissue sampling; cytology and brush, endoscopic biliary drainage without ES, C-tissue sampling; biopsy, nonplacement of

Table 3. Risk Factors of PEP on Univariate Analysis P value Pancreatography first P-tissue sampling; cytology ⫹ brush EBD without ES C-tissue sampling; biopsy Nonplacement of PSDS after ERCP Procedure time: ⱖ30 min IDUS (P-IDUS and C-IDUS) Mean cannulation time P-tissue sampling by any methods P-IDUS Difficulty of cannulation (ⱖ15 min) C-IDUS NOTE. P ⫽ .10 EBD, endoscopic biliary drainage.

⬍.001 ⬍.001 .001 .001 .001 .010 .015 .018 .020 .024 .091 .094

October 2011


Table 4. Risk Factors of PEP on Multivariate Analysis P value Pancreatography first Nonplacement of PSDS after ERCP Procedure time ⱖ30 min P-tissue sampling by any method P-IDUS Difficulty of cannulation (ⱖ15 min)

⬍.001 ⬍.001 ⬍.001 .002 .013 .043

Table 6. The Prevention Effect of PSDS Owing to the Number of Risk Factors Number of risk factors Pancreatitis (⫹) nS group S group Pancreatitis (⫺) nS group S group aP bP

PSDS after ERCP, IDUS (total), mean cannulation time, Ptissue sampling by any method, and P–IDUS. Furthermore, only the significant factors (P ⱖ .10) in the chi-square analysis were included on multivariate (logistic regression) analysis. Table 4 shows the 6 significant risk factors for PEP on multivariate analysis, that is, pancreatography first, nonplacement of PS after ERCP, procedure time of 30 minutes or more, P-tissue sampling by any method, P–IDUS, and difficulty of cannulation (ⱖ15 min).

Relationship Between the Number of Risk Factors for Post–Endoscopic Retrograde Cholangiopancreatography Pancreatitis There was a significant association with PEP and the number of risk factors (P ⫽ .011, Cochran–Armitage test), and there was a statistically significant frequency in PEP when there were more than 3 risk factors (P ⫽ .001) (Table 5). Table 6 also shows the preventive effect of PSDS according to the number of risk factors. There was a nonstatistically significant tendency toward prevention of PEP by the placement of PS in cases showing less than 2 risk factors (P ⫽ .06).

Table 5. Correlation of Number of Risk Factors and PEP

855

ⱕ2

ⱖ3

19 15 4a 224 122 102a

32 16 16b 151 61 90b

⫽ .06. ⫽ .317.

Discussion This RCT showed that placement of PS was feasible and significantly reduced the rate of PEP in patients with risk factors on FAS (P ⫽ .021), although there was no statistically significant difference in the prevention of PEP on ITT analysis (P ⫽ .076). Moreover, it also identified risk factors for PEP. PEP is generally mild and requires only conservative treatment. However, substantial complications sometimes occur, and on occasion they are fatal. Furthermore, the frequency and severity of PEP increases in cases with risk factors. The various mechanisms of PEP include postprocedural impaired drainage of the pancreatic duct owing to papillary edema, or spasm of the sphincter of Oddi, or both.10 –13,15,22,25,26 Other mechanisms included local injury of the papilla and pancreatic duct as a result of the procedure, or forceful and repetitive contrast injections causing local inflammation.1–20,26 This may lead to premature intracellular activation of proteolytic enzymes, consequently causing further damage and local inflammation as

856

SOFUNI ET AL

indicated by increased levels of cytokines, and possible initiation of a systemic inflammatory response with multiorgan involvement.1,10,11,13 To date, several attempts of medication to reduce synthesis and secretion of proteolytic enzymes and prevent PEP have been attempted, including administration of chemoprophylactic and antiproteolytic drugs.9,10 The effects varied and there is no consensus on whether chemoprophylaxis is useful for preventing PEP. We gave a proteolytic enzyme, ulinastatin, according to the guidelines in Japan. The ulinastatin may have influenced the frequency of PEP. We should take that into consideration as a limitation when we evaluate the frequency of PEP in this study. Several mainly retrospective, randomized studies9 –12,14,18 –21 have evaluated endoscopic drainage using PS with flanges on both sides, unlike PSDS used in the present study, for preventing PEP in high-risk patients. The results of our study suggested that the frequency of PEP decreased, and the PS provided a maintained drainage route when the papilla was blocked as a result of edema, or spasm of sphincter of Oddi, or both, after the procedure. Recently, Freeman17 concluded that the insertion of PS for highrisk cases reduced the incidence of PEP by 4%—23%. Moreover, our previous RCT showed that placement of PSDS significantly prevented PEP in all patients who needed ERCP, as the result of a preliminary single-center RCT in 2007.15 However, previous studies essentially consisted of patients at high risk for PEP, and most were retrospective studies. Furthermore, it is necessary to conduct RCTs to evaluate whether placement of PSDS would prevent pancreatitis after ERCP-related procedures for patients with any risk factors of PEP, and to identify risk factors for PEP. We designed the present RCT based on that study. The selection of risk factors and inclusion and exclusion criteria were determined based on various previous studies1–26 and discussions by collaborating organizations. The results of the present study according to the strict ITT analysis showed that the insertion of PSDS often prevented PEP, although there was no statistically significant difference in this prevention (9.4% vs 14.6%; P ⫽ .076). Although FAS analysis showed that the insertion of PSDS significantly reduced the frequency of PEP (7.9% vs 15.2%; P ⫽ .021), the rate of pancreatitis in the nS group was high compared with that of the S group. Drop-out cases should not be excluded from the applicable analytic cases with RCT. Because 37 large institutions participated in the present RCT, ITT analysis was performed with all registered cases, including exclusion cases. Because FAS, except for exclusion cases, reflected the actual clinical situation, the result of FAS have clinical significance. The overall rate of PEP was 11.9% (11.3%, FAS) in all cases with risk factors. This agrees in general with previous reports; the rate of PEP in cases with risk factors ranges from 1% to as high as 40%.1– 8,14 –24,26,29 –32 The registration number of cases per one institution was a mean of 11.5 (range, 2–35) cases. The variation range was owing to the time of the approval by the ethics committee at each facility, which affected the starting time. The mean frequency of PEP in each institution was 11.2 (range, 0%– 60)%, but there was a significant difference in the frequency of PEP among the facilities (P ⫽ .02), which may have influenced the incidence rate of PEP, and statistically therefore is one limitation of this study. Stratification adjustment could have been performed from the statistical point of view. In future studies the number of facilities should be reduced and the number of cases in each facility should be increased to prevent differences among the


facilities. The difference from the results of our previous RCT may be related to this point. Stent placement failed in 25 cases (11.7%). The causes of unsuccessful PS placement in this study included 3 cases of anatomic deformities of the pancreatic duct, 15 cases of inability to cannulate the pancreatic duct, and 7 cases of difficulty in passing the guidewire deeply enough. The cause of inability to cannulate the pancreatic duct was the result of attempts made after endoscopic papillary balloon dilation and ES. The complications included 3 cases of mild pancreatitis (12.0%). Freeman et al1,20 also reported that unsuccessful cases were at higher risk of pancreatitis. The frequency of the failure of PSDS placement (12%) may have been one cause of the difference between our previous study and this ITT analysis. We should therefore consider adapting the PS insertion method in cases that require it, using the method that involves insertion of a guidewire only before the main procedure. Moreover, occasionally it is difficult to place a stent in cases of anatomic deformity. When we encountered such cases, we carefully used the Radifocus guidewire (Terumo, Co, Ltd) or switched to using 0.018or 0.025-inch guidewire, and generally were able to place the stent. These manipulations would have contributed to a higher incidence of PEP in the S group. The frequency of the failure of PSDS placement (12%), the existence of anatomic deformities of the pancreatic duct, and the dispersion of indication cases with risk factors in both groups (ie, “Pancreatography ⱖ2 times, endoscopic pancreatic sphincterotomy, C-IDUS, and procedure time ⱖ30 min) (Supplementary Appendix 3) may be perceived as limitations of this trial. Freeman et al1,2 reported that the high risks for PEP were both patient-related risk factors (ie, young age, female sex, suspected sphincter of Oddi dysfunction), and procedurerelated risk factors (ie, pancreatic duct injection, difficult or impossible cannulation, and precut [access] sphincterotomy). In the present study, the selection of risk factors and setting of inclusion and exclusion criteria were determined upon consideration of various reports1–26,29 –32 and discussions among collaborating organizations concerning previously reported risk factors in the international literature, but were expanded in the present study (eg, C-IDUS). The rate of PEP with patient-related and procedure-related risk factors in both groups is shown in Supplementary Appendix 4. On multivariate analysis, PEP was related significantly to pancreatography first, nonplacement of PS after ERCP, procedure time of 30 minutes or more, P-tissue sampling by any method, P-IDUS, and difficulty of cannulation (ⱖ15 min). Although pancreatitis caused by significant risk factors on multivariate analysis was reduced by PSDS placement, the relationship between pancreatography first and difficulty of cannulation was statistically significant and tended to reduce procedure time. However, there was no difference in P-tissue sampling by any method with P-IDUS. This could represent an important result. This indicates that PEP might be affected by other factors as well as papillary edema, such as the direct burden placed on the pancreatic duct by P-tissue sampling by any method with P-IDUS. IDUS and P-tissue sampling devices rub against the pancreatic duct in cases in which the upper-side PS length does not extend to the target site. This may cause injury of the pancreatic duct per se, or insufficient drainage. Sometimes, we encountered cases in which it was difficult to cannulate the bile duct. The method of P-GW to achieve selec-

October 2011

tive biliary cannulation was selected according to the judgment of each endoscopist. The method of P-GW was performed in 10.8% of all cases. The frequency of using P-GW in cases with PEP was 34.6%, and the frequency of using P-GW in cases without PEP was 7.9%. The frequency of PEP was statistically high in cases using P-GW (P ⬍ .001). This suggests the possibility of direct damage to the pancreatic duct by the guidewire. When the method cannot be performed in ERCP procedures, a softer and more delicate guidewire should be used. The frequency of pancreatic duct strictures in the body and tail in cases of PEP was relatively high (19.2%). The causes of strictures were cancer, cyst, ductitis, and autoimmune pancreatitis. That is because the 3-cm long PS, which does not extend beyond the stenosis, cannot secure pancreatic juice drainage. Therefore, PSDS of the long type beyond the stenosis is desirable in such cases. Difficult cannulation, long procedure time, and unsuccessful placement of the PS can be possible causes of papillary edema. Pancreatography first could be necessitated by many factors such as papillary edema, or chemical reactions in the upper side branch of the pancreatic duct, which could become more serious owing to insufficient drainage. Therefore, it is necessary to place a PS, to help prevent possible risk factors, as shown in this study. The elucidation of the relation of the number of risk factors and PEP is an important result. Freeman et al2 indicated that the cumulative combination of 4 risk factors increased PEP. There was a significant correlation between the number of risk factors and PEP (P ⫽ .011), and there was significant frequency of PEP when patients had more than 3 risk factors (P ⫽ .001). Moreover, the preventive effect of PSDS according to the number of risk factors revealed a tendency toward prevention of PEP by PS in those cases with 2 risk factors or fewer (P ⫽ .06). In cases with more than 3 risk factors, not only patient-related factors but also several procedure-related factors might be associated with pancreatitis. In other words, other factors apart from papillary edema might affect PEP onset. Regarding risk factors for PEP, not only papilledema but also various other factors have been suggested to be associated with procedures, but these results suggested that it might be possible to reduce progression to severe pancreatitis by the placement of PS for papillary edema only. In conclusion, this study showed that the placement of PS significantly reduces PEP according to the FAS and identified risk factors for PEP. There was a statistically significant correlation between the existence of risk factors for PEP and the number of risk factors. Moreover, this study indicated that the onset of PEP was complexly associated with many factors, among which papillary edema was prominent.

Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Clinical Gastroenterology and Hepatology at www.cghjournal.org and doi:10.1016/ j.cgh.2011.06.033. References 1. Freeman ML, Nalini M, Guda M. Prevention of post-ERCP pancreatitis: a comprehensive review. Gastrointest Endosc 2004;59: 845– 864.


857

2. Freeman ML, DiSario JA, Nelson DB, et al. Risk factors for postERCP pancreatitis: a prospective, multicenter study. Gastrointest Endosc 2001;54:425– 434. 3. Cheng CL, Sherman S, Watkins JL, et al. Risk factors for postERCP pancreatitis: a prospective multicenter study. Am J Gastroenterol 2006;101:139 –147. 4. Freeman ML, Nelson DB, Sherman S, et al. Complications of endoscopic biliary sphincterotomy. N Engl J Med 1996;335:909 – 918. 5. Masci E, Toti G, Mariani A, et al. Complications of diagnostic and therapeutic ERCP: a prospective multicenter study. Am J Gastroenterol 2001;96:417– 423. 6. Mehta SN, Pavone E, Barkun JS, et al. Predictors of post-ERCP complications in patients with suspected choledocholithiasis. Endoscopy 1998;30:457– 463. 7. Sherman S, Ruffolo TA, Hawes RH, et al. A prospective series with emphasis on the increased risk associates with sphincter of Oddi dysfunction and nondilated bile ducts. Gastroenterology 1991;101:778 –782. 8. Vandervoort J, Soetikno RM, Tham TC, et al. Risk factors for complications after performance of ERCP. Gastrointest Endosc 2002;56:652– 656. 9. Elton E, Howell DA, Parsons WG, et al. Endoscopic pancreatic sphincterotomy: indications, outcome, and a safe stentless technique. Gastrointest Endosc 1998;47:240 –249. 10. Fazel A, Quadri A, Catalano MF, et al. Does a pancreatic duct stent prevent post-ERCP pancreatitis? A prospective randomized study. Gastrointest Endosc 2003;57:291–294. 11. Tarnasky PR, Palesch YY, Cunningham JT, et al. Pancreatic stenting prevents pancreatitis after biliary sphincterotomy in patients with sphincter of Oddi dysfunction. Gastroenterology 1998;115: 1518 –1524. 12. Smithline A, Silverman W, Rogers D, et al. Effect of prophylactic main pancreatic duct stenting on the incidence of biliary endoscopic sphincterotomy-induced pancreatitis in high-risk patients. Gastrointest Endosc 1993;39:652– 657. 13. Singh P, Das A, Isenberg G, et al. Does prophylactic pancreatic stent placement reduce the risk of post-ERCP acute pancreatitis? A meta-analysis of controlled trials. Gastrointest Endosc 2004;60:544 –550. 14. Catalano MF, Linder JD, Chak A, et al. Endoscopic management of adenoma of the major duodenal papilla. Gastrointest Endosc 2004;59:225–232. 15. Tsuchiya T, Itoi T, Sofuni A, et al. A temporary inner unflanged 5Fr pancreatic duct stent to prevent post-ERCP pancreatitis. A preliminary and single center randomized controlled trial study. J Hepatobiliary Pancreat Surg 2007;14:302–307. 16. Sherman S, Lehman G, Freeman M, et al. Risk factors for postERCP pancreatitis: a prospective multicenter study. Am J Gastroenterol 1997;92(Suppl 9):1639. 17. Freeman ML. Role of pancreatic stents in prevention of postERCP pancreatitis. Pancreas 2004;5:322–327. 18. Fogel EL, Eversman D, Jamidar P, et al. Sphincter of Oddi dysfunction: pancreaticobiliary sphincterotomy with pancreatic stent placement has a lower rate of pancreatitis than biliary sphincterotomy alone. Endoscopy 2002;34:280 –285. 19. Aizawa T, Ueno N. Stent placement in the pancreatic duct prevents pancreatitis after endoscopic sphincter dilation for removal of bile duct stones. Gastrointest Endosc 2001;54:209 –213. 20. Freeman ML, Overby C, Qi D. Pancreatic stent insertion: consequences of failure and results of a modified technique to maximize success. Gastrointest Endosc 2004;59:8 –14. 21. Vandervoort J, Soetikno RM, Montes H, et al. Accuracy and complication rate of brush cytology from bile duct versus pancreatic duct. Gastrointest Endosc 1999;49:322–327. 22. Cotton PB, Lehman G, Vennes J, et al. Endoscopic sphincterot-

858

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

SOFUNI ET AL

omy complications and their management: an attempt at consensus. Gastrointest Endosc 1991;37:383–393. Tarnasky P, Cunningham J, Cotton P, et al. Pancreatic sphincter hypertension increases the risk of post-ERCP pancreatitis. Endoscopy 1997;29:252–257. Sherman S, Troiano FP, Hawes RH, et al. Sphincter of Oddi manometry: decreased risk of clinical pancreatitis with use of a modified aspirating catheter. Gastrointest Endosc 1990;36: 462– 466. Rashdan A, Fogel EL, Mchenry L Jr, et al. Improved stent characteristics for prophylaxis of post-ERCP pancreatitis. Clin Gastroenterol Hepatol 2004;2:322–329. Sofuni A, Maguchi H, Itoi T, et al. Prophylaxis of post-endoscopic retrograde cholangiopancreatography pancreatitis by an endoscopic pancreatic spontaneous dislodgement stent. Clin Gastroenterol Hepatol 2007;5:1339 –1346. Dumonceau JM, Deviere J, Cremer M. A new method of achieving deep cannulation of the common bile duct during endoscopic retrograde cholangiopancreatography. Endoscopy 1998;30:S80. Ito K, Fujita N, Noda Y, et al. Pancreatic guidewire placement for achieving selective biliary cannulation during endoscopic retrograde cholangio-pancreatography. World J Gastroenterol 2008;14:5595– 5600. Mazaki T, Masuda H, Takayama T. Prophylactic pancreatic stent placement and post-ERCP pancreatitis: a systematic review and meta-analysis. Endoscopy 2010;42:842– 853. Dumonceau JM, Andriulli A, Deviere J, et al. European Society of Gastrointestinal Endoscopy. European Society of Gastrointestinal Endoscopy (ESGE) guideline: prophylaxis of post-ERCP pancreatitis. Endoscopy 2010;42:503–515. Chahal P, Tarnasky PR, Petersen BT, et al. Short 5Fr vs long 3Fr pancreatic stents in patients at risk for post-endoscopic retrograde cholangiopancreatography pancreatitis. Clin Gastroenterol Hepatol 2009;7:834 – 839. Testoni PA, Mariani A, Giussani A, et al. Risk factors for postERCP pancreatitis in high- and low-volume centers and among expert and non-expert operators: a prospective multicenter study. Am J Gastroenterol 2010;105:1753–1761.


chi, MD, Department of Gastroenterology, Tokai University; Mitushiro Kida, MD, Department of Gastroenterology, Kitazasto University East Hospital; Takeshi Hisa, MD, Department of Internal Medicine, Saku Central Hospital; Shinji Okaniwa, MD, Department of Gastroenterology, Iida Municipal Hospital; Masataka Kikuyama, MD, Department of Gastroenterology, Hamamatsu Rosai Hospital; Ichiro Yasuda, MD, Department of Gastroenterology, Gifu University; Tsuyoshi Mukai, MD, Department of Gastroenterology, Gifu Municipal Hospital; Masamiichi Enya, MD, Department of Gastroenterology, Takayama Red Cross Hospital; Shinsuke Tada, MD, Department of Gastroenterology, Kyoto University; Yoshihiro Okabe, MD, Department of Gastroenterology, Osaka, Red Cross Hospital; Hidekazu Mukai, MD, Department of Gastroenterology, Yodogawa Christian Hospital; Yoshifumi Arisaka, MD, Department of Gastroenterology, Osaka Medical University; Masayuki Kitano, MD, Department of Gastroenterology and Hepatology, Kinki University; Shujiro Yasumi, MD, Department of Gastroenterology, Kitano Hospital; Hiromu Kutsumi, MD, Department of Gastroenterology, Kobe University; Keiji Hanada, MD, Center for Gastroendoscopy, Onomichi General Hospital; Shomei Ryozawa, MD, Department of Gastroenterology and Hepatology, Yamaguchi University Graduate School of Medicine; Koji Yoshida, MD, Department of Hepatology and Pancreatlogy, Kawasaki Medical School; Hirofumi Kawamoto, MD, Department of Gastroenterology and Hepatology, Okayama University; Toru Ueki, MD, Department of Gastroenterology, Hiroshima City Hospital/Fukuyama City Hospital; Tamito Sasaki, MD, Department of Gastroenterology and Hepatology, Hiroshima University; Yoshiya Furukawa, MD, Department of Gastroenterology, Hiroshima Red Cross Hospital and Atomic Bomb Survivors Hospital; Seisuke Okamura, MD, Department of Gastroenterology and Oncology, Tokushima University; Yoshinobu Okabe, MD, Department of Gastroenterology, Kurume University; and Toshi- haru Ueki, MD, Department of Gastroenterology, Fukuoka University Chikushi Hospital.

Appendix 1. The Japan Pancreatic Stent-Study Group consists of Hiroyuki Maguchi, MD, and Kuniyuki Takahashi, MD, Center for Gastroenterology, Teine-Keijinkai Hospital; Hiroyuki Hisai, MD, Department of Gastroenterology, Date Red Cross Hospital; Atsushi Chiba, MD, and Tsuneshi Fujii, MD, Department of Gastroenterology, Asahikawa City Hospital; Hiroshi Kawakami, MD, Masaki Kuwatani, MD, Kazunori Eto, MD, Shin Haba, MD, Department of Gastroenterology, Hokkaido University Graduate School of Medicine; Tsuyoshi Hayashi, MD, First Department of Internal Medicine, Sapporo Medical University; Nobuyuki Yanagawa, MD, Department of Gastroenterology, Asahikawa-Kosei General Hospital; Atsushi Irisawa, MD, Preparatory office for Aizu Medical Center, Fukushima Medical University; Atsushi Sofuni, MD, and Takao Itoi, MD, Department of Gastroenterology and Hepatology, Tokyo Medical University Hospital; Ikuo Haruyama, MD, Department of Gastroenterology, Toda Chuo General Hospital; Teitetsu Niido, MD, Department of Gastroenterology, Tokyo Medical University Hachioji Medical Center; Yasuharu Yamaguchi, MD, Department of Gastroenterology, Kyorin University; Kensuke Kubota, MD, Department of Gastroenterology, Yokohama City University; Tetsuya Mine, MD, and Yoshiaki Kawagu-

Reprint requests Address requests for reprints to: Atsushi Sofuni, MD, PhD, Department of Gastroenterology and Hepatology, Tokyo Medical University, 6-7-1 Nishishinjuku, Shinjuku-ku, Tokyo 160-0023, Japan. e-mail: [email protected]; fax: (81) 3-5381-6654. Acknowledgments The authors thank the members of the Japan Pancreatic StentStudy Group (Appendix 1), Professor Martin Freeman for supporting this study, and Professor Toshio Morizane, MD, Kanagawa Dental College, who specializes in biostatistics and who provided independent statistical analysis of this study. The authors are indebted to Mr Roderick J. Turner and Professor J. Patrick Barron of the Department of International Medical Communications of Tokyo Medical University for their review of this manuscript. All authors were members of the Japan Pancreatic Stent-Study Group. The trial is registered at UMIN-CRT, number UMIN000004324. The full trial protocol can be accessed at http://www.umin.ac.jp/ctr/. Conﬂicts of interest The authors disclose no conﬂicts.