ORIGINAL ARTICLE
ERCP-related adverse events in patients with primary sclerosing cholangitis Udayakumar Navaneethan, MD, Ramprasad Jegadeesan, MD, Shishira Nayak, MD, Vennisvasanth Lourdusamy, MD, Madhusudhan R. Sanaka, MD, John J. Vargo, MD, MPH, Mansour A. Parsi, MD, MPH Cleveland, Ohio, USA
Background: ERCP is frequently used in patients with primary sclerosing cholangitis (PSC) for cancer surveillance and treatment of dominant strictures. Objective: To evaluate the prevalence and risk factors for ERCP-related adverse events in patients with PSC. Design: Retrospective analysis of ERCPs performed from 1998 to 2012. Setting: Referral center. Patients: A total of 294 consecutive patients with PSC who underwent a total of 657 ERCPs. Interventions: ERCP. Main Outcome Measurements: ERCP-related adverse events and predictive factors were determined by univariate and multivariate analyses. Results: ERCP use in patients with PSC showed a significant increase during the second half of the study period (2006-2012) compared with the first half (1998-2005) (437 vs 220 procedures; P Z .04). Primary cannulation was successful in 634 procedures (96.6%) or in 271 of 294 patients (92.2%). Access to the bile duct was achieved with a needle-knife in 19 procedures (2.9%), whereas ERCP was unsuccessful in 4 of 657 procedures (0.6%), and successful percutaneous drainage was performed. Post-ERCP pancreatitis (PEP) was diagnosed in 8 (1.2%), cholangitis in 16 (2.4%), and bleeding in 4 (0.7%) procedures. Overall, risk of any adverse event was 28 of 657 (4.3%) procedures. On multivariate analysis, performing biliary sphincterotomy (odds ratio [OR] 5.04; 95% confidence interval [CI], 2.01-12.60; P Z .001) and passage of a guidewire into the pancreatic duct (OR 4.54; 95% CI, 1.44-14.30; P Z .010) were independently associated with an increased risk of any adverse event. Limitations: Retrospective study. Conclusion: Cholangitis appears to be the most common adverse event despite intraprocedural antibiotic use. There was a low risk of adverse events in patients with PSC undergoing ERCP. (Gastrointest Endosc 2014;-:1-10.)
Primary sclerosing cholangitis (PSC) is a chronic, cholestatic disease usually seen in association with underlying inflammatory bowel disease.1-3 ERCP often is warranted in a patient with PSC when there is clinical deterioration with worsening pruritus, jaundice, or bacterial cholangitis
in order to detect a dominant stricture.4 Dominant stricture develops in 45% to 58% of patients during the disease course with PSC and is characterized by stenosis %1.5 mm in the common bile duct or %1 mm in the hepatic ducts.5 Treatment of a dominant stricture with balloon dilation
Abbreviations: CCA, cholangiocarcinoma; FISH, fluorescence in situ hybridization; PEP, post-ERCP pancreatitis; PSC, primary sclerosing cholangitis.
Received May 12, 2014. Accepted June 11, 2014.
DISCLOSURE: J. Vargo is a consultant for Olympus, Boston Scientific, Cook Endoscopy, and Ethicon Endosurgery. All other authors disclosed no financial relationships relevant to this article. Copyright ª 2014 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2014.06.030
www.giejournal.org
Current affiliations: Department of Gastroenterology, Digestive Disease Institute, Cleveland Clinic, Cleveland, Ohio, USA. Reprint requests: Mansour A. Parsi, MD, MPH, Section for Advanced Endoscopy and Pancreatobiliary Disorders, Digestive Disease Institute, Desk A30, The Cleveland Clinic, 9500 Euclid Ave, Cleveland, OH 44195. If you would like to chat with an author of this article, you may contact Dr Parsi at [email protected].
Volume
-,
No.
-
: 2014 GASTROINTESTINAL ENDOSCOPY 1
ERCP-related adverse events and PSC
Navaneethan et al
alone or balloon dilation with stent placement is attempted to provide biliary drainage and to reduce symptoms.5 In the presence of a dominant stricture, the suspicion for cholangiocarcinoma (CCA) is high, and their distinction is often difficult. Although the advent of magnetic resonance cholangiography has largely replaced diagnostic ERCP, ERCP still may be required to diagnose early intrahepatic PSC.6 In addition, ERCP is the preferred method of biliary drainage and diagnosis of early CCA, so that appropriate treatment can be instituted. Brush cytology during ERCP, in combination with fluorescence in situ hybridization (FISH), have become the standard diagnostic methods for detection of CCA.5,7,8 Because patients with PSC undergo repeated ERCP examinations and have stenotic and narrow bile ducts, they may be at risk for development of ERCP-related adverse events, in particular cholangitis. Previous studies have shown that patients with PSC have a slightly higher incidence of ERCP-related adverse events compared with the population without PSC.9-13 In a large series on the safety of ERCP in patients from Europe with PSC, the overall risk of adverse events was 9%.14 However, there is a lack of large studies from North America assessing ERCP-related adverse events in patients with PSC and factors determining it, so that appropriate measures can be instituted in the prevention of these adverse events. We have a large cohort of patients with PSC treated and followed-up in our institution, and our aim was to analyze the trend of use of ERCPs in our institution, the overall prevalence of adverse events, and the factors predicting the risk of adverse events in patients with PSC.
METHODS Patients The Cleveland Clinic electronic medical records database was queried to identify patients with PSC who had undergone ERCP procedures from 1998 to 2012. The medical records of all identified patients were manually reviewed. Demographic, clinical, and procedural data and data regarding adverse events were collected. The study was approved by the Cleveland Clinic Institutional Review Board. Inclusion criteria were patients aged O18 years and those with PSC who underwent ERCP. Exclusion criteria were liver transplant recipient status and surgically altered anatomy (hepaticojejunostomy).
ERCP procedures Our usual protocol is to accomplish wire-guided cannulation of the bile duct by using a sphincterotome and guidewire or extraction balloon and guidewire. Needleknife sphincterotomy was used if access could not be obtained by standard cannulation equipment. Inadvertent cannulation or injection of the pancreatic duct was 2 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2014
Take-home Message Cholangitis appears to be the most common adverse event in patients with primary sclerosing cholangitis (PSC) undergoing ERCP, in spite of antibiotic use. There is a low risk of adverse events in patients with PSC undergoing ERCP. Biliary sphincterotomy and accidental wire passes into the pancreatic duct increase the risk of adverse events.
followed by placement of a prophylactic pancreatic stent to reduce the risk of pancreatitis. Use of rectal indomethacin was instituted only after the study period and was not used during this study to prevent PEP. Our usual protocol of managing dominant strictures was to dilate the stricture followed by short-term stenting (2 weeks) in these patients. However, in some instances with severe PSC and a dominant stricture, balloon dilation alone was performed. In patients without a dominant stricture, if there was inadequate drainage of contrast material, short-term stenting (2 weeks) was performed. Patients were sent home with 5 days of antibiotics after the procedure to prevent cholangitis. Brush cytology specimens were collected for cytology and/or FISH if a dominant stricture was encountered, and these were obtained before dilation. FISH was used only during the last year of the study period (2012). Currently, in our institution, 2 sets of brushings are obtained (1 for cytology, 1 for FISH) for any patient with PSC who has a dominant stricture at the time of ERCP. All patients received prophylactic intravenous antibiotics during the procedure. ERCP procedures were performed by 1 of 6 experienced interventional endoscopists who used standard endoscopic equipment. Data on therapeutic procedures during ERCP such as sphincterotomy (both biliary and pancreatic), the presence of a dominant stricture, stricture dilation, biliary and pancreatic stent placement, biliary stone extraction, and the need for percutaneous transhepatic cholangiography were collected.
Adverse events Post-ERCP adverse events were defined based on the American Society for Gastrointestinal Endoscopy workshop.15 Adverse events were recorded up to 30 days after the procedures. PEP was defined as new-onset abdominal pain and more than 3-fold elevation in serum amylase and/or lipase levels for up to 2 weeks after the procedure. Cholangitis was defined as the presence of fever or leukocytosis or positive blood cultures requiring intravenous antibiotics within 14 days after ERCP. Bleeding was defined as clinical evidence of hemorrhage, with a decrease in hemoglobin O2 g/dL and/or the need for endoscopic or other methods of obtaining hemostasis within 14 days after ERCP. Perforation was defined as extravasation www.giejournal.org
Navaneethan et al
of contrast material on imaging after the procedure. In addition, 30-day adverse events were recorded in these patients. The timing of the adverse event after the procedure, need for admission and/or readmission to the hospital and the length of hospital stay were recorded for all adverse events occurring up to 30 days after the procedure.
Outcome measures Our primary study outcome was to analyze the trends in utilization of ERCP in PSC patients and to evaluate the frequency and risk factors predicting the 30-day adverse events after the ERCP.
Statistical analysis Descriptive statistics were computed for all factors. These include medians, 25th and 75th percentiles, range or mean and standard deviation for continuous factors, and frequencies and percentages for categorical factors. Wilcoxon rank sum tests for continuous factors and Pearson chi-square or the Fisher exact tests for categorical factors were used. Logistic regression models were constructed by including variables that had significant univariate associations with post-ERCP adverse events and then performing backward stepwise selection with a removal criterion of P O .05. R 2.10.1 software (The R Foundation for Statistical Computing, Vienna, Austria) was used to perform all analyses.
RESULTS During the study period (between 1998 and 2012), 294 consecutive patients with PSC underwent 657 ERCPs. In the study cohort, ERCP use continued to increase during the latter half of the study period (2006-2012) compared with the first half of the study period (1998-2005) (437 vs 220 procedures; P Z .04). The median age of the patients was 47 years (range 12-85 years), and 31.0% were female. Primary cannulation was successful in 634 procedures (96.6%). In the group with failed primary cannulation, access to the bile duct was achieved with a needle-knife in 19 procedures (2.9%). ERCP was unsuccessful in 4 of 657 procedures (0.6%), and successful percutaneous transhepatic cholangiography was performed. Overall, primary wire-guided or contrast-material injection for cannulation was successful in 271 of 294 (92.2%) procedures. In patients with a native ampulla without a prior biliary sphincterotomy, cannulation was successful in 240 of 263 (91.3%) procedures. None of the patients had liver transplantation performed within the 30-day follow-up period. Biliary sphincterotomy was performed at ERCP in 31 of 294 patients (10.5%). Sixty-four patients (21.8%) had a prior biliary sphincterotomy during previous procedures www.giejournal.org
ERCP-related adverse events and PSC
from outside institutions. Pancreatic sphincterotomy was performed in 3 of 294 patients (1.0%). Biliary sphincterotomy or extension of prior biliary sphincterotomy for stone removal or other therapeutic procedures was performed in 64 procedures (9.7%), whereas biliary sphincterotomy was done previously in 279 procedures (42.5%). Dominant strictures were found during 180 procedures. Overall, balloon dilation alone was performed in 36 procedures, balloon dilation with stenting in 99 procedures, and short-term biliary stenting alone in 45 procedures. There was no difference in overall adverse events between various groups (data not shown). Bile duct brushing was done in 244 of 657 (37.1%) procedures, biliary stents were placed in 244 of 657 (37.1%) procedures, and stone removal was performed in 99 of 657 (15.0%) procedures. Extraction balloons were used for stone removal in 80 procedures, whereas a Dormia basket and balloon was used in the remaining 19 procedures. A total of 54 patients underwent liver transplantation on follow-up. However, none of the patients had liver transplantation performed within the 30-day follow-up period. Among the 294 patients, 34 patients underwent repeat ERCP for elevated bilirubin levels in the absence of cholangitis within the 30-day follow-up period. These patients had initial ERCPs for evaluation of dominant strictures. Among the 34 patients, dominant strictures were seen in 7 patients. Patients with dominant strictures had balloon dilation with stent placement. Patients without dominant strictures also underwent stent placement because of inadequate drainage of contrast material at the end of the procedure. Patients who did not have improvement in bilirubin levels had the stents removed during follow-up ERCP. These patients had liver cirrhosis resulting in elevated bilirubin levels and severe disease not amenable to endoscopic intervention.
Univariate analyses Evaluated risk factors for 30-day post-ERCP adverse events are presented in Table 1. PEP was diagnosed in 8 (1.2%), cholangitis in 16 (2.4%), and bleeding in 4 (0.7%) procedures as shown in Figure 1. Thus, the overall risk of the presence of any adverse event was 28 of 657 (4.3%) procedures. Factors that increased the risk of any adverse event were pancreatic sphincterotomy, pancreatic duct cannulation (O0), contrast material injection into the pancreatic duct, biliary sphincterotomy, precut with a needle-knife, balloon dilation of the stricture, brushings of the stricture, and biliary stent placement (Table 1). The risk of adverse events increased with multiple procedures including balloon dilation for a biliary stricture, brushings of the biliary stricture for cancer surveillance, and stent placement. The timing of adverse events after ERCP was between 1 and 5.5 days, and median length of stay after admission for any adverse events was 4.0 days. Multiple (O1) stents were placed in 15 patients who Volume
-,
No.
-
: 2014 GASTROINTESTINAL ENDOSCOPY 3
ERCP-related adverse events and PSC
Navaneethan et al
TABLE 1. Risk factors for post-ERCP adverse events in 294 patients undergoing 657 ERCP procedures for primary sclerosing cholangitis No. of adverse events/no. of procedures (%) 28/657 (4.3%)
Odds ratio (95% confidence interval)
P value
Female
10/205 (4.9)
0.81 (0.37-1.78)
.60
Male
18/452 (4.0)
0.88 (0.77-1.01)
.07
24.4 (3.9)
0.73 (0.49-1.09)
.13
Yes
4/19 (21.1)
5.65 (1.77-17.98)
.011
No
24/638 (3.8)
12.00 (2.10-68.52)
.024
8.04 (3.13-20.67)
! .001
2.65 (1.09-6.46)
.036
7.03 (3.13-15.79)
! .001
Variable Sex
Age, median (range) With adverse event
44.5 (29.5-55.8)
Without adverse event
48.0 (38.8-59.0)
BMI, mean (SD) Precut with needle-knife
Pancreatic sphincterotomy Yes
2/6 (33.3)
No
26/651 (4.0)
One or more accidental wire passes into pancreatic duct Yes
7/32 (21.9)
No
21/625 (3.4)
Contrast material injection into pancreatic duct Yes
7/77 (9.1)
No
21/580 (3.6)
Biliary sphincterotomy Yes
11/64 (17.2)
No
17/593 (2.9)
Previous biliary sphincterotomy Yes
10/279 (3.6)
No
18/378 (4.8)
! .001
Balloon dilation Yes
12/165 (7.3)
No
16/492 (3.3)
2.33 (1.08-5.03)
.032
3.20 (1.45-7.05)
.004
4.51 (1.95-10.40)
! .001
Brushings Yes
18/244 (7.4)
No
10/413 (2.4)
Biliary stent placement Yes
20/244 (8.2)
No
8/413 (1.9)
4 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2014
www.giejournal.org
Navaneethan et al
ERCP-related adverse events and PSC
TABLE 1. Continued No. of adverse events/no. of procedures (%) 28/657 (4.3%)
Odds ratio (95% confidence interval)
P value
Yes
4/99 (4.0)
0.94 (0.32-2.76)
1.00
No
24/558 (4.3) 5.35 (2.64-10.84)
.99
Variable Stone removal
Timing of adverse event since ERCP, median (IQR)
1 (1.0-5.5)
Length of stay for readmission, median (IQR)
4.0 (2-10)
.99
BMI, Body mass index; SD, standard deviation; IQR, interquartile range.
30-day post ERCP adverse events
Number of adverse events
18 16 14 12 10 8 6 4 2 0
Post-ERCP pancreatitis
Cholangitis
Bleeding
Type of adverse event Figure 1. Box plot of the various adverse events after ERCP in patients with primary sclerosing cholangitis.
had 44 ERCPs performed in our cohort of patients. The overall adverse event rate in this group was 2 of 44 (4.5%), which was no different from the group without multiple stents. Figure 2 demonstrates the presence of a dominant stricture and balloon dilation of the stricture at the time of ERCP.
Adverse events Among the 4 patients with bleeding, 3 had bleeding during the procedure, and 1 patient had bleeding 48 hours after the procedure. Two patients with intraprocedural bleeding were admitted to the hospital for observation, and these patients stayed overnight without the need for any blood transfusion. They were discharged the next day in stable condition. The other patient had self-limited bleeding during the procedure and was discharged home on the same day of the procedure. The patient with bleeding 48 hours after the procedure presented with melena and a hemoglobin drop of O2 g/dL. The patient was admitted to the intensive care unit for 3 days and required www.giejournal.org
transfusion of 4 units of blood and an urgent endoscopy with epinephrine injection and placement of 2 hemoclips to attain hemostasis. The patient was discharged in stable condition after 8 days of hospital stay. Thus, 3 adverse events were classified as mild and 1 as a moderate severity adverse event. Among the 16 patients with cholangitis, 7 developed cholangitis within the first 48 hours, 4 patients developed cholangitis within 2 to 7 days, and 5 patients developed cholangitis O7 days after ERCP. All patients with cholangitis were treated with intravenous antibiotics. In addition to antibiotics, a subsequent ERCP with a therapeutic procedure such as balloon dilation, stent replacement, or stone retrieval was required by 9 patients. None of the patients with cholangitis required admission to the intensive care unit. There was no mortality among the patients who had adverse events after the procedure. Thus, 7 patients were classified as having a mild adverse event, whereas 9 patients had a moderate severity adverse event. Evaluated risk factors for PEP are presented in Table 2. On univariate analysis, factors that increased the risk of PEP were pancreatic sphincterotomy, cannulation of the pancreatic duct at least once, contrast material injection into the pancreatic duct, biliary sphincterotomy, and precut biliary sphincterotomy. All 8 patients with PEP were admitted to the hospital and were managed with supportive care including intravenous fluids, analgesics, and antibiotics if clinically indicated. The mean timing of PEP after ERCP was 1.3 days, and the median length of stay after admission for PEP was 3.5 days. None of the patients had severe pancreatitis or required admission to the intensive care unit. Seven of the 8 patients had mild severity adverse events, whereas 1 had a moderate severity adverse event. The adverse events were classified according to the lexicon of adverse events after ERCP and are summarized in Table 3.
Multivariate analyses All clinically relevant variables with P values ! .10 in the univariate logistic regression analyses were selected as Volume
-,
No.
-
: 2014 GASTROINTESTINAL ENDOSCOPY 5
ERCP-related adverse events and PSC
Navaneethan et al
Figure 2. A, A dominant stricture in the left intrahepatic ducts. B, Dilation of dominant stricture with a balloon.
candidates for the multivariate model building. These factors were pancreatic sphincterotomy, guidewire cannulation of the pancreatic duct more than once, contrast material injection into the pancreatic duct, precut with a needle-knife, and biliary sphincterotomy. Biliary sphincterotomy (odds ratio [OR] 5.04; 95% confidence interval [CI], 2.01-12.6; P Z .001) and accidental wire passes into the pancreatic duct at least once (OR 4.54; 95% CI, 1.44-14.3; P Z .010) independently increased the risk of any adverse event, whereas pancreatic sphincterotomy (OR 1.45; 95% CI, 0.21-9.95; P Z .70), precut with a needle-knife (OR 1.56; 95% CI, 0.36-6.75; P Z .55), and contrast material injection into the pancreatic duct (OR 0.63; 95% CI, 0.16-2.51; P Z .55) did not attain statistical significance (Table 4).
DISCUSSION PSC is a chronic, progressive, inflammatory disorder of intrahepatic and extrahepatic bile ducts. The disease course of PSC is characterized by the development of clinically significant biliary strictures, dominant strictures, and cholangitis and/or stone formation necessitating multiple endoscopic biliary interventions and ultimately resulting in progressive liver cirrhosis requiring liver transplantation. Use of endoscopic interventions may improve survival and postpone the need for liver transplantation.16-18 We analyzed the trends in the use of ERCP in patients with PSC in our institution and evaluated the frequency and risk factors predicting the 30-day adverse events after ERCP in 294 patients with PSC who underwent 657 ERCPs. We observed an increase in the use of ERCP in patients with PSC. This is likely because of an increase in referrals for complex ERCP procedures to tertiary-care referral centers.19 Although there has been an increase in the number of referrals and complexity of procedures, there was no significant difference in adverse events 6 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2014
from 1998 to 2005 (3.2%) and 2006 to 2012 (6.1%; P Z .17). In one of the largest series of ERCPs in patients with PSC from Europe, the success rate in primary cannulation of the biliary tree was 88.2%.14 In our study population, successful primary cannulation was achieved in most patients, and ERCP was unsuccessful in only 4 of 657 (0.6%) procedures in patients who underwent a successful percutaneous approach for biliary drainage. Overall, primary cannulation was successful in 271 of 294 procedures (92.2%). This may be attributed to enhanced experience in doing ERCPs by interventional endoscopists who perform larger volumes of ERCP procedures (O1500 ERCPs/year as a group). In general, the adverse event rate for ERCP is reported to be 3% to 11%.9-11 However, among patients with PSC, the overall incidence of adverse events has been reported to be between 7.3% and 18%.12-14 Nonetheless, in our series, the overall adverse event rate was 4.3%. The previously identified risk factors for developing any adverse event are multiple interventions during ERCP, prolonged duration of the procedure,13,14 and the procedure performed in an acute setting like cholangitis.18 In our study, when primary wire-guided or contrast-material guided cannulation failed, use of adjunct procedures such as precut biliary sphincterotomy and pancreatic sphincterotomy were all significantly associated with higher adverse event rates on univariate analysis. Also, accidental wire passage into the pancreatic duct increased the chances of an adverse event to a level similar to findings of a previous study.14 We had a limited number of patients in our study with PEP, and we did not encounter any sidebranch perforations based on retrospective review of records. Biliary sphincterotomy is a well-known risk factor that increases the risk of post-ERCP adverse events by at least 2-fold.20 We also observed that the adverse event rate increased. As expected, previous biliary sphincterotomy www.giejournal.org
Navaneethan et al
ERCP-related adverse events and PSC
TABLE 2. Risk factors for PEP in 294 patients undergoing 657 ERCP procedures for primary sclerosing cholangitis No. of PEP cases/no. of procedures (%) 8/657 (1.2%)
Odds ratio (95% confidence interval)
P value
Female
4/205 (2.0)
0.45 (0.11-1.81)
.26
Male
4/452 (0.9)
0.77 (0.60-1.00)
.05
23.17 (1.99)
0.56 (0.25-1.23)
.15
Yes
2/19 (10.5)
10.47 (1.99-55.11)
.027
No
6/638 (0.9)
53.76 (8.20-351.05)
.002
22.14 (5.26-93.15)
! .001
4.65 (1.09-19.84)
.04
9.82 (2.39-40.25)
.004
Variable Sex
Age, median (range) With adverse event
37.5 (26.8-45.0)
Without adverse event
47.5 (38.0-59.0)
BMI, mean (SD) Precut with needle-knife
Pancreatic sphincterotomy Yes
2/6 (33.3)
No
6/651 (0.92)
One or more accidental wire passes into pancreatic duct Yes
4/32 (12.5)
No
4/625 (0.64)
Contrast material injection into pancreatic duct Yes
3/77 (3.9)
No
5/580 (0.8)
Biliary sphincterotomy Yes
4/64 (6.2)
No
4/593 (0.67)
Previous biliary sphincterotomy Yes
2/279 (0.71)
No
6/378 (1.58)
.03
Balloon dilation Yes
4/165 (2.4)
No
4/492 (0.81)
3.02 (0.75-12.23)
.011
2.85 (0.68-12.04)
.16
2.85 (0.68-12.04)
.16
Brushings Yes
5/244 (2.0)
No
3/413 (0.73)
Biliary stent placement Yes
5/244 (2.0)
No
3/413 (0.73)
Timing of PEP since ERCP, mean (SD)
1.25 (1.98)
1.22 (1.00-1.47)
.05
Length of stay for readmission, median (IQR)
3.5 (3-6.25)
1.28 (1.15-1.43)
! .001
PEP, Post-ERCP pancreatitis; BMI, body mass index; SD, standard deviation; IQR, interquartile range.
www.giejournal.org
Volume
-,
No.
-
: 2014 GASTROINTESTINAL ENDOSCOPY 7
ERCP-related adverse events and PSC
Navaneethan et al
TABLE 3. Classification of post-ERCP adverse events* Adverse events (severity)
Mild
Moderate
Bleeding, definite (N Z 4)
3, During procedure
1, After procedure
7, None required ICU admission, and all patients required !3 d hospitalization
9, Repeat procedure (ERCP) for the adverse event
7, O3 d hospitalization, not requiring ICU admission
1, Between 3-10 d hospitalization and not requiring ICU admission
Cholangitis, probable (N Z 16) N Z 7 (within 48 h) N Z 4 (2-7 d after procedure) N Z 5 (7-14 d after procedure) Pancreatitis, definite (N Z 8) N Z 8 (!14 d after procedure) ICU, Intensive care unit. *There were no severe or fatal adverse events.
TABLE 4. Multivariate analysis of risk factors of post-ERCP adverse events Odds ratio (95% CI)
P value
Biliary sphincterotomy
5.04 (2.01-12.6)
.001
One or more accidental passes into pancreatic duct
4.54 (1.44-14.3)
.01
Pancreatic sphincterotomy
1.45 (0.21-9.95)
.70
Contrast material injection into pancreatic duct
0.63 (0.16-2.51)
.51
Precut with needle-knife
1.56 (0.36-6.75)
.55
Variable
CI, Confidence interval.
decreased the risk of any adverse event after ERCP on univariate analysis, also similar to findings of a previous study.14 This finding is in line with findings of the study from Finland, where investigators observed that previous papillotomy decreased the risk of PEP by facilitating primary cannulation and preventing slippage of contrast material or guidewire into the pancreatic duct.14 A third of patients in our cohort had dominant strictures. Therapeutic procedures were required more often in these patients, and the presence of a dominant biliary stricture was significantly associated with higher adverse event rates. This could be attributed to the increased need for intrabiliary interventions like balloon dilation and brushing, which by itself independently increases the risk of adverse events. In our study, we observed that use of biliary stenting to provide adequate biliary drainage was associated with higher adverse event rates, especially cholangitis, on univariate analysis. We did not find an association between stone removal and the development of any adverse event. Because pancreatitis is the most common adverse event after ERCP, much effort had been underway to delineate 8 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2014
the factors predicting it. In the largest series of ERCP in patients with PSC from Finland, the rate of PEP was 9.0%, which is significantly higher than the rate of 1.2% reported in our series.14 The likely explanation for this could be the routine use of pancreatic duct stenting after cannulation or injection of the pancreatic duct in our institution throughout the study period. Younger patient age at the time of ERCP marginally increased the risk of PEP in our study, as observed in previously published studies.21-23 Because patients with PSC are young, especially during disease diagnosis, greater caution needs to be exercised in this group of patients. Female sex is a risk factor for developing an adverse event after ERCP, including PEP.14,21 However, in our subset of patients with PSC who underwent ERCP, female sex was not found to be a contributory factor, similar to findings of a previously published study.24 Pancreatic sphincterotomy increased the propensity to develop PEP. The risk of PEP increased with passage of a guidewire into the pancreatic duct.25 Contrast material injection into the PD is a well-known risk factor for PEP,11,26-28 a finding observed in our study as well. www.giejournal.org
Navaneethan et al
Although all of these factors were significant on univariate analysis, the factors became insignificant on multivariate analysis because of low number of adverse events. The rate of cholangitis in our patient population was 2.1%, which was higher than the rate of PEP. Patients with PSC are at high risk of cholangitis because contrast material injection and inadequate drainage increase the risk. The presence of dominant strictures and biliary sphincterotomy increased the risk of cholangitis in our study. Among biliary interventions, only brushing and stent placement augmented the risk of cholangitis. Also we observed the lower the diameter of the stent used, the higher the adverse event of cholangitis. However, we did not have enough adverse events to detect the impact of each of these individual factors on adverse events on multivariate analysis. Smaller-diameter stents (7F) are sometime required in patients with PSC because these patients have thin, narrow-caliber bile ducts, and it would be very difficult to place a 10F stent. In patients with dominant strictures, placement of two 7F stents or one 7F and one 10F stent was attempted. We did not have enough observed cholangitis events to determine the impact of stent size on adverse events. Based on the observed percentage of events in our study, it would require at least 1200 ERCPs to develop a model with 5 factors to study the impact of stent size. A multicenter study would be the only way to answer this question. There were several limitations in our study, including its retrospective design and a heterogeneous patient population undergoing ERCP procedures over a 15-year period. In addition, the number of adverse events was low, limiting the power of the study. The major strength of this study was the inclusion of all ERCPs performed for a large cohort of patients with PSC and the inclusion of all the techniques at ERCP, allowing a rigorous analysis of adverse events. We reported the 30-day adverse event rate in this study. It was not a routine protocol to do a 30-day telephone call to the patients who underwent ERCP during the study period. However, because all patients with PSC had centralized follow-up at the Cleveland Clinic because of tertiary-level care required in these patients, all of these patients had a follow-up with a physician or a nurse practitioner within the 30 days and were asked about any issues after the ERCP procedure. So, we were able to capture 30-day adverse events in all of the patients included in our study. In case patients presented to outside centers, they were all referred to the Cleveland Clinic, and, hence, we are confident that we captured all the adverse events in the study. In this study, which to our knowledge has the largest sample size of ERCPs performed in patients with PSC, cholangitis appears to be the most common adverse event. Overall, the risk of adverse events was very low. We also identified several procedural factors that could be implemented in clinical practice to see whether there is a www.giejournal.org
ERCP-related adverse events and PSC
decrease in adverse events after ERCP procedures in patients with PSC. To conclude, this study reinforces the safety of ERCP as a diagnostic and therapeutic tool for detecting biliary dysplasia and to manage biliary stricture and its adverse events in patients with PSC.
ACKNOWLEDGEMENTS We would like to thank Jeffrey Hammel for help with statistical analysis.
REFERENCES 1. Wiesner RH, Grambsch PM, Dickson ER, et al. Primary sclerosing cholangitis: natural history, prognostic factors and survival analysis. Hepatology 1989;10:430-6. 2. Faubion WA, Loftus EV, Sandborn WJ, et al. Pediatric “PSC-IBD”: a descriptive report of associated inflammatory bowel disease among pediatric patients with PSC. J Pediatr Gastroenterol Nutrit 2001;33: 296-300. 3. Olsson R, Danielsson A, Jarnerot G, et al. Prevalence of primary sclerosing cholangitis in patients with ulcerative colitis. Gastroenterology 1991;100:1319-23. 4. Singh S, Talwalkar JA. Primary sclerosing cholangitis: diagnosis, prognosis, and management. Clin Gastroenterol Hepatol 2013;11:898-907. 5. Chapman R, Fevery J, Kalloo A, et al. Diagnosis and management of primary sclerosing cholangitis. Hepatology 2010;51:660-78. 6. Berstad AE, Aabakken L, Smith HJ, et al. Diagnostic accuracy of magnetic resonance and endoscopic retrograde cholangiography in primary sclerosing cholangitis. Clin Gastroenterol Hepatol 2006;4:514-20. 7. Navaneethan U, Njei B, Venkatesh PG, et al. Fluorescence in situ hybridization for diagnosis of cholangiocarcinoma in primary sclerosing cholangitis: a systematic review and meta-analysis. Gastrointest Endosc. Epub 2013 Dec 19. 8. Trikudanathan G, Navaneethan U, Njei B, et al. Diagnostic yield of bile duct brushings for cholangiocarcinoma in primary sclerosing cholangitis: a systematic review and meta-analysis. Gastrointest Endosc 2014;79:783-9. 9. Bilbao MK, Dotter CT, Lee TG, et al. Adverse events of endoscopic retrograde cholangiopancreatography (ERCP). A study of 10,000 cases. Gastroenterology 1976;70:314-20. 10. Silviera ML, Seamon MJ, Porshinsky B, et al. Adverse events related to endoscopic retrograde cholangiopancreatography: a comprehensive clinical review. J Gastrointes Liver Dis 2009;18:73-82. 11. Vandervoort J, Soetikno RM, Tham TC, et al. Risk factors for adverse events after performance of ERCP. Gastrointest Endosc 2002;56:652-6. 12. Enns R, Eloubeidi MA, Mergener K, et al. Predictors of successful clinical and laboratory outcomes in patients with primary sclerosing cholangitis undergoing endoscopic retrograde cholangiopancreatography. Can J Gastroenterol 2003;17:243-8. 13. Bangarulingam SY, Gossard AA, Petersen BT, et al. Adverse events of endoscopic retrograde cholangiopancreatography in primary sclerosing cholangitis. Am J Gastroenterol 2009;104:855-60. 14. Ismail S, Kylänpää L, Mustonen H, et al. Risk factors for complications of ERCP in primary sclerosing cholangitis. Endoscopy 2012;44:1133-8. 15. Cotton PB, Eisen GM, Aabakken L, et al. A lexicon for endoscopic adverse events: report of an ASGE workshop. Gastrointest Endosc 2010;71:446-54. 16. Gluck M, Cantone NR, Brandabur JJ, et al. A twenty-year experience with endoscopic therapy for symptomatic primary sclerosing cholangitis. J Clin Gastroenterol 2008;42:1032-9.
Volume
-,
No.
-
: 2014 GASTROINTESTINAL ENDOSCOPY 9
ERCP-related adverse events and PSC
Navaneethan et al
17. Baluyut AR, Sherman S, Lehman GA, et al. Impact of endoscopic therapy on the survival of patients with primary sclerosing cholangitis. Gastrointest Endosc 2001;53:308-12. 18. van den Hazel SJ, Wolfhagen EH, van Buuren HR, et al. Prospective risk assessment of endoscopic retrograde cholangiography in patients with primary sclerosing cholangitis. Dutch PSC Study Group. Endoscopy 2000;32:779-82. 19. James PD, Kaplan GG, Myers RP, et al. Decreasing mortality from acute biliary diseases that require endoscopic retrograde cholangiopancreatography: a nationwide cohort study. Clin Gastroenterol Hepatol. Epub 2013 Oct 2. 20. Dumonceau JM, Andriulli A, Deviere J, et al. European Society of Gastrointestinal Endoscopy (ESGE) Guideline: prophylaxis of postERCP pancreatitis. Endoscopy 2010;42:503-15. 21. Etzel JP, Eng SC, Ko CW, et al. Adverse events after ERCP inpatients with primary sclerosing cholangitis. Gastrointest Endosc 2008;67:643-8. 22. Freeman ML, DiSario JA, Nelson DB, et al. Risk factors for post-ERCP pancreatitis: a prospective, multicenter study. Gastrointest Endosc 2001;54:425-34.
10 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2014
23. Wang P, Li ZS, Liu F, et al. Risk factors for ERCP-related complications: a prospective multicenter study. Am J Gastroenterol 2009;104: 31-40. 24. Alkhatib, Hilden K, Adler DG. Comorbidities, sphincterotomy, and balloon dilation predict post-ERCP adverse events in PSC patients: operator experience is protective. Dig Dis Sci 2011;56: 3685-8. 25. Lella F, Bagnolo F, Colombo E, et al. A simple way of avoiding post-ERCP pancreatitis. Gastrointest Endosc 2004;59:830-4. 26. Cotton PB, Garrow DA, Gallagher J, et al. Risk factors for complications after ERCP: a multivariate analysis of 11,497 procedures over 12 years. Gastrointest Endosc 2009;70:80-8. 27. Testoni PA, Mariani A, Giussani A, et al. Risk factors for post-ERCP pancreatitis in high and low-volume centers and among expert and non-expert operators: a prospective multicenter study. Am J Gastroenterol 2010;105:1753-61. 28. Halttunen J, Keränen I, Udd M, et al. Pancreatic sphincterotomy versus needle knife precut in difficult biliary cannulation. Surg Endosc 2009;23:745-9.
www.giejournal.org
ERCP-related adverse events in patients with primary sclerosing cholangitis Udayakumar Navaneethan, MD, Ramprasad Jegadeesan, MD, Shishira Nayak, MD, Vennisvasanth Lourdusamy, MD, Madhusudhan R. Sanaka, MD, John J. Vargo, MD, MPH, Mansour A. Parsi, MD, MPH Cleveland, Ohio, USA
Background: ERCP is frequently used in patients with primary sclerosing cholangitis (PSC) for cancer surveillance and treatment of dominant strictures. Objective: To evaluate the prevalence and risk factors for ERCP-related adverse events in patients with PSC. Design: Retrospective analysis of ERCPs performed from 1998 to 2012. Setting: Referral center. Patients: A total of 294 consecutive patients with PSC who underwent a total of 657 ERCPs. Interventions: ERCP. Main Outcome Measurements: ERCP-related adverse events and predictive factors were determined by univariate and multivariate analyses. Results: ERCP use in patients with PSC showed a significant increase during the second half of the study period (2006-2012) compared with the first half (1998-2005) (437 vs 220 procedures; P Z .04). Primary cannulation was successful in 634 procedures (96.6%) or in 271 of 294 patients (92.2%). Access to the bile duct was achieved with a needle-knife in 19 procedures (2.9%), whereas ERCP was unsuccessful in 4 of 657 procedures (0.6%), and successful percutaneous drainage was performed. Post-ERCP pancreatitis (PEP) was diagnosed in 8 (1.2%), cholangitis in 16 (2.4%), and bleeding in 4 (0.7%) procedures. Overall, risk of any adverse event was 28 of 657 (4.3%) procedures. On multivariate analysis, performing biliary sphincterotomy (odds ratio [OR] 5.04; 95% confidence interval [CI], 2.01-12.60; P Z .001) and passage of a guidewire into the pancreatic duct (OR 4.54; 95% CI, 1.44-14.30; P Z .010) were independently associated with an increased risk of any adverse event. Limitations: Retrospective study. Conclusion: Cholangitis appears to be the most common adverse event despite intraprocedural antibiotic use. There was a low risk of adverse events in patients with PSC undergoing ERCP. (Gastrointest Endosc 2014;-:1-10.)
Primary sclerosing cholangitis (PSC) is a chronic, cholestatic disease usually seen in association with underlying inflammatory bowel disease.1-3 ERCP often is warranted in a patient with PSC when there is clinical deterioration with worsening pruritus, jaundice, or bacterial cholangitis
in order to detect a dominant stricture.4 Dominant stricture develops in 45% to 58% of patients during the disease course with PSC and is characterized by stenosis %1.5 mm in the common bile duct or %1 mm in the hepatic ducts.5 Treatment of a dominant stricture with balloon dilation
Abbreviations: CCA, cholangiocarcinoma; FISH, fluorescence in situ hybridization; PEP, post-ERCP pancreatitis; PSC, primary sclerosing cholangitis.
Received May 12, 2014. Accepted June 11, 2014.
DISCLOSURE: J. Vargo is a consultant for Olympus, Boston Scientific, Cook Endoscopy, and Ethicon Endosurgery. All other authors disclosed no financial relationships relevant to this article. Copyright ª 2014 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2014.06.030
www.giejournal.org
Current affiliations: Department of Gastroenterology, Digestive Disease Institute, Cleveland Clinic, Cleveland, Ohio, USA. Reprint requests: Mansour A. Parsi, MD, MPH, Section for Advanced Endoscopy and Pancreatobiliary Disorders, Digestive Disease Institute, Desk A30, The Cleveland Clinic, 9500 Euclid Ave, Cleveland, OH 44195. If you would like to chat with an author of this article, you may contact Dr Parsi at [email protected].
Volume
-,
No.
-
: 2014 GASTROINTESTINAL ENDOSCOPY 1
ERCP-related adverse events and PSC
Navaneethan et al
alone or balloon dilation with stent placement is attempted to provide biliary drainage and to reduce symptoms.5 In the presence of a dominant stricture, the suspicion for cholangiocarcinoma (CCA) is high, and their distinction is often difficult. Although the advent of magnetic resonance cholangiography has largely replaced diagnostic ERCP, ERCP still may be required to diagnose early intrahepatic PSC.6 In addition, ERCP is the preferred method of biliary drainage and diagnosis of early CCA, so that appropriate treatment can be instituted. Brush cytology during ERCP, in combination with fluorescence in situ hybridization (FISH), have become the standard diagnostic methods for detection of CCA.5,7,8 Because patients with PSC undergo repeated ERCP examinations and have stenotic and narrow bile ducts, they may be at risk for development of ERCP-related adverse events, in particular cholangitis. Previous studies have shown that patients with PSC have a slightly higher incidence of ERCP-related adverse events compared with the population without PSC.9-13 In a large series on the safety of ERCP in patients from Europe with PSC, the overall risk of adverse events was 9%.14 However, there is a lack of large studies from North America assessing ERCP-related adverse events in patients with PSC and factors determining it, so that appropriate measures can be instituted in the prevention of these adverse events. We have a large cohort of patients with PSC treated and followed-up in our institution, and our aim was to analyze the trend of use of ERCPs in our institution, the overall prevalence of adverse events, and the factors predicting the risk of adverse events in patients with PSC.
METHODS Patients The Cleveland Clinic electronic medical records database was queried to identify patients with PSC who had undergone ERCP procedures from 1998 to 2012. The medical records of all identified patients were manually reviewed. Demographic, clinical, and procedural data and data regarding adverse events were collected. The study was approved by the Cleveland Clinic Institutional Review Board. Inclusion criteria were patients aged O18 years and those with PSC who underwent ERCP. Exclusion criteria were liver transplant recipient status and surgically altered anatomy (hepaticojejunostomy).
ERCP procedures Our usual protocol is to accomplish wire-guided cannulation of the bile duct by using a sphincterotome and guidewire or extraction balloon and guidewire. Needleknife sphincterotomy was used if access could not be obtained by standard cannulation equipment. Inadvertent cannulation or injection of the pancreatic duct was 2 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2014
Take-home Message Cholangitis appears to be the most common adverse event in patients with primary sclerosing cholangitis (PSC) undergoing ERCP, in spite of antibiotic use. There is a low risk of adverse events in patients with PSC undergoing ERCP. Biliary sphincterotomy and accidental wire passes into the pancreatic duct increase the risk of adverse events.
followed by placement of a prophylactic pancreatic stent to reduce the risk of pancreatitis. Use of rectal indomethacin was instituted only after the study period and was not used during this study to prevent PEP. Our usual protocol of managing dominant strictures was to dilate the stricture followed by short-term stenting (2 weeks) in these patients. However, in some instances with severe PSC and a dominant stricture, balloon dilation alone was performed. In patients without a dominant stricture, if there was inadequate drainage of contrast material, short-term stenting (2 weeks) was performed. Patients were sent home with 5 days of antibiotics after the procedure to prevent cholangitis. Brush cytology specimens were collected for cytology and/or FISH if a dominant stricture was encountered, and these were obtained before dilation. FISH was used only during the last year of the study period (2012). Currently, in our institution, 2 sets of brushings are obtained (1 for cytology, 1 for FISH) for any patient with PSC who has a dominant stricture at the time of ERCP. All patients received prophylactic intravenous antibiotics during the procedure. ERCP procedures were performed by 1 of 6 experienced interventional endoscopists who used standard endoscopic equipment. Data on therapeutic procedures during ERCP such as sphincterotomy (both biliary and pancreatic), the presence of a dominant stricture, stricture dilation, biliary and pancreatic stent placement, biliary stone extraction, and the need for percutaneous transhepatic cholangiography were collected.
Adverse events Post-ERCP adverse events were defined based on the American Society for Gastrointestinal Endoscopy workshop.15 Adverse events were recorded up to 30 days after the procedures. PEP was defined as new-onset abdominal pain and more than 3-fold elevation in serum amylase and/or lipase levels for up to 2 weeks after the procedure. Cholangitis was defined as the presence of fever or leukocytosis or positive blood cultures requiring intravenous antibiotics within 14 days after ERCP. Bleeding was defined as clinical evidence of hemorrhage, with a decrease in hemoglobin O2 g/dL and/or the need for endoscopic or other methods of obtaining hemostasis within 14 days after ERCP. Perforation was defined as extravasation www.giejournal.org
Navaneethan et al
of contrast material on imaging after the procedure. In addition, 30-day adverse events were recorded in these patients. The timing of the adverse event after the procedure, need for admission and/or readmission to the hospital and the length of hospital stay were recorded for all adverse events occurring up to 30 days after the procedure.
Outcome measures Our primary study outcome was to analyze the trends in utilization of ERCP in PSC patients and to evaluate the frequency and risk factors predicting the 30-day adverse events after the ERCP.
Statistical analysis Descriptive statistics were computed for all factors. These include medians, 25th and 75th percentiles, range or mean and standard deviation for continuous factors, and frequencies and percentages for categorical factors. Wilcoxon rank sum tests for continuous factors and Pearson chi-square or the Fisher exact tests for categorical factors were used. Logistic regression models were constructed by including variables that had significant univariate associations with post-ERCP adverse events and then performing backward stepwise selection with a removal criterion of P O .05. R 2.10.1 software (The R Foundation for Statistical Computing, Vienna, Austria) was used to perform all analyses.
RESULTS During the study period (between 1998 and 2012), 294 consecutive patients with PSC underwent 657 ERCPs. In the study cohort, ERCP use continued to increase during the latter half of the study period (2006-2012) compared with the first half of the study period (1998-2005) (437 vs 220 procedures; P Z .04). The median age of the patients was 47 years (range 12-85 years), and 31.0% were female. Primary cannulation was successful in 634 procedures (96.6%). In the group with failed primary cannulation, access to the bile duct was achieved with a needle-knife in 19 procedures (2.9%). ERCP was unsuccessful in 4 of 657 procedures (0.6%), and successful percutaneous transhepatic cholangiography was performed. Overall, primary wire-guided or contrast-material injection for cannulation was successful in 271 of 294 (92.2%) procedures. In patients with a native ampulla without a prior biliary sphincterotomy, cannulation was successful in 240 of 263 (91.3%) procedures. None of the patients had liver transplantation performed within the 30-day follow-up period. Biliary sphincterotomy was performed at ERCP in 31 of 294 patients (10.5%). Sixty-four patients (21.8%) had a prior biliary sphincterotomy during previous procedures www.giejournal.org
ERCP-related adverse events and PSC
from outside institutions. Pancreatic sphincterotomy was performed in 3 of 294 patients (1.0%). Biliary sphincterotomy or extension of prior biliary sphincterotomy for stone removal or other therapeutic procedures was performed in 64 procedures (9.7%), whereas biliary sphincterotomy was done previously in 279 procedures (42.5%). Dominant strictures were found during 180 procedures. Overall, balloon dilation alone was performed in 36 procedures, balloon dilation with stenting in 99 procedures, and short-term biliary stenting alone in 45 procedures. There was no difference in overall adverse events between various groups (data not shown). Bile duct brushing was done in 244 of 657 (37.1%) procedures, biliary stents were placed in 244 of 657 (37.1%) procedures, and stone removal was performed in 99 of 657 (15.0%) procedures. Extraction balloons were used for stone removal in 80 procedures, whereas a Dormia basket and balloon was used in the remaining 19 procedures. A total of 54 patients underwent liver transplantation on follow-up. However, none of the patients had liver transplantation performed within the 30-day follow-up period. Among the 294 patients, 34 patients underwent repeat ERCP for elevated bilirubin levels in the absence of cholangitis within the 30-day follow-up period. These patients had initial ERCPs for evaluation of dominant strictures. Among the 34 patients, dominant strictures were seen in 7 patients. Patients with dominant strictures had balloon dilation with stent placement. Patients without dominant strictures also underwent stent placement because of inadequate drainage of contrast material at the end of the procedure. Patients who did not have improvement in bilirubin levels had the stents removed during follow-up ERCP. These patients had liver cirrhosis resulting in elevated bilirubin levels and severe disease not amenable to endoscopic intervention.
Univariate analyses Evaluated risk factors for 30-day post-ERCP adverse events are presented in Table 1. PEP was diagnosed in 8 (1.2%), cholangitis in 16 (2.4%), and bleeding in 4 (0.7%) procedures as shown in Figure 1. Thus, the overall risk of the presence of any adverse event was 28 of 657 (4.3%) procedures. Factors that increased the risk of any adverse event were pancreatic sphincterotomy, pancreatic duct cannulation (O0), contrast material injection into the pancreatic duct, biliary sphincterotomy, precut with a needle-knife, balloon dilation of the stricture, brushings of the stricture, and biliary stent placement (Table 1). The risk of adverse events increased with multiple procedures including balloon dilation for a biliary stricture, brushings of the biliary stricture for cancer surveillance, and stent placement. The timing of adverse events after ERCP was between 1 and 5.5 days, and median length of stay after admission for any adverse events was 4.0 days. Multiple (O1) stents were placed in 15 patients who Volume
-,
No.
-
: 2014 GASTROINTESTINAL ENDOSCOPY 3
ERCP-related adverse events and PSC
Navaneethan et al
TABLE 1. Risk factors for post-ERCP adverse events in 294 patients undergoing 657 ERCP procedures for primary sclerosing cholangitis No. of adverse events/no. of procedures (%) 28/657 (4.3%)
Odds ratio (95% confidence interval)
P value
Female
10/205 (4.9)
0.81 (0.37-1.78)
.60
Male
18/452 (4.0)
0.88 (0.77-1.01)
.07
24.4 (3.9)
0.73 (0.49-1.09)
.13
Yes
4/19 (21.1)
5.65 (1.77-17.98)
.011
No
24/638 (3.8)
12.00 (2.10-68.52)
.024
8.04 (3.13-20.67)
! .001
2.65 (1.09-6.46)
.036
7.03 (3.13-15.79)
! .001
Variable Sex
Age, median (range) With adverse event
44.5 (29.5-55.8)
Without adverse event
48.0 (38.8-59.0)
BMI, mean (SD) Precut with needle-knife
Pancreatic sphincterotomy Yes
2/6 (33.3)
No
26/651 (4.0)
One or more accidental wire passes into pancreatic duct Yes
7/32 (21.9)
No
21/625 (3.4)
Contrast material injection into pancreatic duct Yes
7/77 (9.1)
No
21/580 (3.6)
Biliary sphincterotomy Yes
11/64 (17.2)
No
17/593 (2.9)
Previous biliary sphincterotomy Yes
10/279 (3.6)
No
18/378 (4.8)
! .001
Balloon dilation Yes
12/165 (7.3)
No
16/492 (3.3)
2.33 (1.08-5.03)
.032
3.20 (1.45-7.05)
.004
4.51 (1.95-10.40)
! .001
Brushings Yes
18/244 (7.4)
No
10/413 (2.4)
Biliary stent placement Yes
20/244 (8.2)
No
8/413 (1.9)
4 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2014
www.giejournal.org
Navaneethan et al
ERCP-related adverse events and PSC
TABLE 1. Continued No. of adverse events/no. of procedures (%) 28/657 (4.3%)
Odds ratio (95% confidence interval)
P value
Yes
4/99 (4.0)
0.94 (0.32-2.76)
1.00
No
24/558 (4.3) 5.35 (2.64-10.84)
.99
Variable Stone removal
Timing of adverse event since ERCP, median (IQR)
1 (1.0-5.5)
Length of stay for readmission, median (IQR)
4.0 (2-10)
.99
BMI, Body mass index; SD, standard deviation; IQR, interquartile range.
30-day post ERCP adverse events
Number of adverse events
18 16 14 12 10 8 6 4 2 0
Post-ERCP pancreatitis
Cholangitis
Bleeding
Type of adverse event Figure 1. Box plot of the various adverse events after ERCP in patients with primary sclerosing cholangitis.
had 44 ERCPs performed in our cohort of patients. The overall adverse event rate in this group was 2 of 44 (4.5%), which was no different from the group without multiple stents. Figure 2 demonstrates the presence of a dominant stricture and balloon dilation of the stricture at the time of ERCP.
Adverse events Among the 4 patients with bleeding, 3 had bleeding during the procedure, and 1 patient had bleeding 48 hours after the procedure. Two patients with intraprocedural bleeding were admitted to the hospital for observation, and these patients stayed overnight without the need for any blood transfusion. They were discharged the next day in stable condition. The other patient had self-limited bleeding during the procedure and was discharged home on the same day of the procedure. The patient with bleeding 48 hours after the procedure presented with melena and a hemoglobin drop of O2 g/dL. The patient was admitted to the intensive care unit for 3 days and required www.giejournal.org
transfusion of 4 units of blood and an urgent endoscopy with epinephrine injection and placement of 2 hemoclips to attain hemostasis. The patient was discharged in stable condition after 8 days of hospital stay. Thus, 3 adverse events were classified as mild and 1 as a moderate severity adverse event. Among the 16 patients with cholangitis, 7 developed cholangitis within the first 48 hours, 4 patients developed cholangitis within 2 to 7 days, and 5 patients developed cholangitis O7 days after ERCP. All patients with cholangitis were treated with intravenous antibiotics. In addition to antibiotics, a subsequent ERCP with a therapeutic procedure such as balloon dilation, stent replacement, or stone retrieval was required by 9 patients. None of the patients with cholangitis required admission to the intensive care unit. There was no mortality among the patients who had adverse events after the procedure. Thus, 7 patients were classified as having a mild adverse event, whereas 9 patients had a moderate severity adverse event. Evaluated risk factors for PEP are presented in Table 2. On univariate analysis, factors that increased the risk of PEP were pancreatic sphincterotomy, cannulation of the pancreatic duct at least once, contrast material injection into the pancreatic duct, biliary sphincterotomy, and precut biliary sphincterotomy. All 8 patients with PEP were admitted to the hospital and were managed with supportive care including intravenous fluids, analgesics, and antibiotics if clinically indicated. The mean timing of PEP after ERCP was 1.3 days, and the median length of stay after admission for PEP was 3.5 days. None of the patients had severe pancreatitis or required admission to the intensive care unit. Seven of the 8 patients had mild severity adverse events, whereas 1 had a moderate severity adverse event. The adverse events were classified according to the lexicon of adverse events after ERCP and are summarized in Table 3.
Multivariate analyses All clinically relevant variables with P values ! .10 in the univariate logistic regression analyses were selected as Volume
-,
No.
-
: 2014 GASTROINTESTINAL ENDOSCOPY 5
ERCP-related adverse events and PSC
Navaneethan et al
Figure 2. A, A dominant stricture in the left intrahepatic ducts. B, Dilation of dominant stricture with a balloon.
candidates for the multivariate model building. These factors were pancreatic sphincterotomy, guidewire cannulation of the pancreatic duct more than once, contrast material injection into the pancreatic duct, precut with a needle-knife, and biliary sphincterotomy. Biliary sphincterotomy (odds ratio [OR] 5.04; 95% confidence interval [CI], 2.01-12.6; P Z .001) and accidental wire passes into the pancreatic duct at least once (OR 4.54; 95% CI, 1.44-14.3; P Z .010) independently increased the risk of any adverse event, whereas pancreatic sphincterotomy (OR 1.45; 95% CI, 0.21-9.95; P Z .70), precut with a needle-knife (OR 1.56; 95% CI, 0.36-6.75; P Z .55), and contrast material injection into the pancreatic duct (OR 0.63; 95% CI, 0.16-2.51; P Z .55) did not attain statistical significance (Table 4).
DISCUSSION PSC is a chronic, progressive, inflammatory disorder of intrahepatic and extrahepatic bile ducts. The disease course of PSC is characterized by the development of clinically significant biliary strictures, dominant strictures, and cholangitis and/or stone formation necessitating multiple endoscopic biliary interventions and ultimately resulting in progressive liver cirrhosis requiring liver transplantation. Use of endoscopic interventions may improve survival and postpone the need for liver transplantation.16-18 We analyzed the trends in the use of ERCP in patients with PSC in our institution and evaluated the frequency and risk factors predicting the 30-day adverse events after ERCP in 294 patients with PSC who underwent 657 ERCPs. We observed an increase in the use of ERCP in patients with PSC. This is likely because of an increase in referrals for complex ERCP procedures to tertiary-care referral centers.19 Although there has been an increase in the number of referrals and complexity of procedures, there was no significant difference in adverse events 6 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2014
from 1998 to 2005 (3.2%) and 2006 to 2012 (6.1%; P Z .17). In one of the largest series of ERCPs in patients with PSC from Europe, the success rate in primary cannulation of the biliary tree was 88.2%.14 In our study population, successful primary cannulation was achieved in most patients, and ERCP was unsuccessful in only 4 of 657 (0.6%) procedures in patients who underwent a successful percutaneous approach for biliary drainage. Overall, primary cannulation was successful in 271 of 294 procedures (92.2%). This may be attributed to enhanced experience in doing ERCPs by interventional endoscopists who perform larger volumes of ERCP procedures (O1500 ERCPs/year as a group). In general, the adverse event rate for ERCP is reported to be 3% to 11%.9-11 However, among patients with PSC, the overall incidence of adverse events has been reported to be between 7.3% and 18%.12-14 Nonetheless, in our series, the overall adverse event rate was 4.3%. The previously identified risk factors for developing any adverse event are multiple interventions during ERCP, prolonged duration of the procedure,13,14 and the procedure performed in an acute setting like cholangitis.18 In our study, when primary wire-guided or contrast-material guided cannulation failed, use of adjunct procedures such as precut biliary sphincterotomy and pancreatic sphincterotomy were all significantly associated with higher adverse event rates on univariate analysis. Also, accidental wire passage into the pancreatic duct increased the chances of an adverse event to a level similar to findings of a previous study.14 We had a limited number of patients in our study with PEP, and we did not encounter any sidebranch perforations based on retrospective review of records. Biliary sphincterotomy is a well-known risk factor that increases the risk of post-ERCP adverse events by at least 2-fold.20 We also observed that the adverse event rate increased. As expected, previous biliary sphincterotomy www.giejournal.org
Navaneethan et al
ERCP-related adverse events and PSC
TABLE 2. Risk factors for PEP in 294 patients undergoing 657 ERCP procedures for primary sclerosing cholangitis No. of PEP cases/no. of procedures (%) 8/657 (1.2%)
Odds ratio (95% confidence interval)
P value
Female
4/205 (2.0)
0.45 (0.11-1.81)
.26
Male
4/452 (0.9)
0.77 (0.60-1.00)
.05
23.17 (1.99)
0.56 (0.25-1.23)
.15
Yes
2/19 (10.5)
10.47 (1.99-55.11)
.027
No
6/638 (0.9)
53.76 (8.20-351.05)
.002
22.14 (5.26-93.15)
! .001
4.65 (1.09-19.84)
.04
9.82 (2.39-40.25)
.004
Variable Sex
Age, median (range) With adverse event
37.5 (26.8-45.0)
Without adverse event
47.5 (38.0-59.0)
BMI, mean (SD) Precut with needle-knife
Pancreatic sphincterotomy Yes
2/6 (33.3)
No
6/651 (0.92)
One or more accidental wire passes into pancreatic duct Yes
4/32 (12.5)
No
4/625 (0.64)
Contrast material injection into pancreatic duct Yes
3/77 (3.9)
No
5/580 (0.8)
Biliary sphincterotomy Yes
4/64 (6.2)
No
4/593 (0.67)
Previous biliary sphincterotomy Yes
2/279 (0.71)
No
6/378 (1.58)
.03
Balloon dilation Yes
4/165 (2.4)
No
4/492 (0.81)
3.02 (0.75-12.23)
.011
2.85 (0.68-12.04)
.16
2.85 (0.68-12.04)
.16
Brushings Yes
5/244 (2.0)
No
3/413 (0.73)
Biliary stent placement Yes
5/244 (2.0)
No
3/413 (0.73)
Timing of PEP since ERCP, mean (SD)
1.25 (1.98)
1.22 (1.00-1.47)
.05
Length of stay for readmission, median (IQR)
3.5 (3-6.25)
1.28 (1.15-1.43)
! .001
PEP, Post-ERCP pancreatitis; BMI, body mass index; SD, standard deviation; IQR, interquartile range.
www.giejournal.org
Volume
-,
No.
-
: 2014 GASTROINTESTINAL ENDOSCOPY 7
ERCP-related adverse events and PSC
Navaneethan et al
TABLE 3. Classification of post-ERCP adverse events* Adverse events (severity)
Mild
Moderate
Bleeding, definite (N Z 4)
3, During procedure
1, After procedure
7, None required ICU admission, and all patients required !3 d hospitalization
9, Repeat procedure (ERCP) for the adverse event
7, O3 d hospitalization, not requiring ICU admission
1, Between 3-10 d hospitalization and not requiring ICU admission
Cholangitis, probable (N Z 16) N Z 7 (within 48 h) N Z 4 (2-7 d after procedure) N Z 5 (7-14 d after procedure) Pancreatitis, definite (N Z 8) N Z 8 (!14 d after procedure) ICU, Intensive care unit. *There were no severe or fatal adverse events.
TABLE 4. Multivariate analysis of risk factors of post-ERCP adverse events Odds ratio (95% CI)
P value
Biliary sphincterotomy
5.04 (2.01-12.6)
.001
One or more accidental passes into pancreatic duct
4.54 (1.44-14.3)
.01
Pancreatic sphincterotomy
1.45 (0.21-9.95)
.70
Contrast material injection into pancreatic duct
0.63 (0.16-2.51)
.51
Precut with needle-knife
1.56 (0.36-6.75)
.55
Variable
CI, Confidence interval.
decreased the risk of any adverse event after ERCP on univariate analysis, also similar to findings of a previous study.14 This finding is in line with findings of the study from Finland, where investigators observed that previous papillotomy decreased the risk of PEP by facilitating primary cannulation and preventing slippage of contrast material or guidewire into the pancreatic duct.14 A third of patients in our cohort had dominant strictures. Therapeutic procedures were required more often in these patients, and the presence of a dominant biliary stricture was significantly associated with higher adverse event rates. This could be attributed to the increased need for intrabiliary interventions like balloon dilation and brushing, which by itself independently increases the risk of adverse events. In our study, we observed that use of biliary stenting to provide adequate biliary drainage was associated with higher adverse event rates, especially cholangitis, on univariate analysis. We did not find an association between stone removal and the development of any adverse event. Because pancreatitis is the most common adverse event after ERCP, much effort had been underway to delineate 8 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2014
the factors predicting it. In the largest series of ERCP in patients with PSC from Finland, the rate of PEP was 9.0%, which is significantly higher than the rate of 1.2% reported in our series.14 The likely explanation for this could be the routine use of pancreatic duct stenting after cannulation or injection of the pancreatic duct in our institution throughout the study period. Younger patient age at the time of ERCP marginally increased the risk of PEP in our study, as observed in previously published studies.21-23 Because patients with PSC are young, especially during disease diagnosis, greater caution needs to be exercised in this group of patients. Female sex is a risk factor for developing an adverse event after ERCP, including PEP.14,21 However, in our subset of patients with PSC who underwent ERCP, female sex was not found to be a contributory factor, similar to findings of a previously published study.24 Pancreatic sphincterotomy increased the propensity to develop PEP. The risk of PEP increased with passage of a guidewire into the pancreatic duct.25 Contrast material injection into the PD is a well-known risk factor for PEP,11,26-28 a finding observed in our study as well. www.giejournal.org
Navaneethan et al
Although all of these factors were significant on univariate analysis, the factors became insignificant on multivariate analysis because of low number of adverse events. The rate of cholangitis in our patient population was 2.1%, which was higher than the rate of PEP. Patients with PSC are at high risk of cholangitis because contrast material injection and inadequate drainage increase the risk. The presence of dominant strictures and biliary sphincterotomy increased the risk of cholangitis in our study. Among biliary interventions, only brushing and stent placement augmented the risk of cholangitis. Also we observed the lower the diameter of the stent used, the higher the adverse event of cholangitis. However, we did not have enough adverse events to detect the impact of each of these individual factors on adverse events on multivariate analysis. Smaller-diameter stents (7F) are sometime required in patients with PSC because these patients have thin, narrow-caliber bile ducts, and it would be very difficult to place a 10F stent. In patients with dominant strictures, placement of two 7F stents or one 7F and one 10F stent was attempted. We did not have enough observed cholangitis events to determine the impact of stent size on adverse events. Based on the observed percentage of events in our study, it would require at least 1200 ERCPs to develop a model with 5 factors to study the impact of stent size. A multicenter study would be the only way to answer this question. There were several limitations in our study, including its retrospective design and a heterogeneous patient population undergoing ERCP procedures over a 15-year period. In addition, the number of adverse events was low, limiting the power of the study. The major strength of this study was the inclusion of all ERCPs performed for a large cohort of patients with PSC and the inclusion of all the techniques at ERCP, allowing a rigorous analysis of adverse events. We reported the 30-day adverse event rate in this study. It was not a routine protocol to do a 30-day telephone call to the patients who underwent ERCP during the study period. However, because all patients with PSC had centralized follow-up at the Cleveland Clinic because of tertiary-level care required in these patients, all of these patients had a follow-up with a physician or a nurse practitioner within the 30 days and were asked about any issues after the ERCP procedure. So, we were able to capture 30-day adverse events in all of the patients included in our study. In case patients presented to outside centers, they were all referred to the Cleveland Clinic, and, hence, we are confident that we captured all the adverse events in the study. In this study, which to our knowledge has the largest sample size of ERCPs performed in patients with PSC, cholangitis appears to be the most common adverse event. Overall, the risk of adverse events was very low. We also identified several procedural factors that could be implemented in clinical practice to see whether there is a www.giejournal.org
ERCP-related adverse events and PSC
decrease in adverse events after ERCP procedures in patients with PSC. To conclude, this study reinforces the safety of ERCP as a diagnostic and therapeutic tool for detecting biliary dysplasia and to manage biliary stricture and its adverse events in patients with PSC.
ACKNOWLEDGEMENTS We would like to thank Jeffrey Hammel for help with statistical analysis.
REFERENCES 1. Wiesner RH, Grambsch PM, Dickson ER, et al. Primary sclerosing cholangitis: natural history, prognostic factors and survival analysis. Hepatology 1989;10:430-6. 2. Faubion WA, Loftus EV, Sandborn WJ, et al. Pediatric “PSC-IBD”: a descriptive report of associated inflammatory bowel disease among pediatric patients with PSC. J Pediatr Gastroenterol Nutrit 2001;33: 296-300. 3. Olsson R, Danielsson A, Jarnerot G, et al. Prevalence of primary sclerosing cholangitis in patients with ulcerative colitis. Gastroenterology 1991;100:1319-23. 4. Singh S, Talwalkar JA. Primary sclerosing cholangitis: diagnosis, prognosis, and management. Clin Gastroenterol Hepatol 2013;11:898-907. 5. Chapman R, Fevery J, Kalloo A, et al. Diagnosis and management of primary sclerosing cholangitis. Hepatology 2010;51:660-78. 6. Berstad AE, Aabakken L, Smith HJ, et al. Diagnostic accuracy of magnetic resonance and endoscopic retrograde cholangiography in primary sclerosing cholangitis. Clin Gastroenterol Hepatol 2006;4:514-20. 7. Navaneethan U, Njei B, Venkatesh PG, et al. Fluorescence in situ hybridization for diagnosis of cholangiocarcinoma in primary sclerosing cholangitis: a systematic review and meta-analysis. Gastrointest Endosc. Epub 2013 Dec 19. 8. Trikudanathan G, Navaneethan U, Njei B, et al. Diagnostic yield of bile duct brushings for cholangiocarcinoma in primary sclerosing cholangitis: a systematic review and meta-analysis. Gastrointest Endosc 2014;79:783-9. 9. Bilbao MK, Dotter CT, Lee TG, et al. Adverse events of endoscopic retrograde cholangiopancreatography (ERCP). A study of 10,000 cases. Gastroenterology 1976;70:314-20. 10. Silviera ML, Seamon MJ, Porshinsky B, et al. Adverse events related to endoscopic retrograde cholangiopancreatography: a comprehensive clinical review. J Gastrointes Liver Dis 2009;18:73-82. 11. Vandervoort J, Soetikno RM, Tham TC, et al. Risk factors for adverse events after performance of ERCP. Gastrointest Endosc 2002;56:652-6. 12. Enns R, Eloubeidi MA, Mergener K, et al. Predictors of successful clinical and laboratory outcomes in patients with primary sclerosing cholangitis undergoing endoscopic retrograde cholangiopancreatography. Can J Gastroenterol 2003;17:243-8. 13. Bangarulingam SY, Gossard AA, Petersen BT, et al. Adverse events of endoscopic retrograde cholangiopancreatography in primary sclerosing cholangitis. Am J Gastroenterol 2009;104:855-60. 14. Ismail S, Kylänpää L, Mustonen H, et al. Risk factors for complications of ERCP in primary sclerosing cholangitis. Endoscopy 2012;44:1133-8. 15. Cotton PB, Eisen GM, Aabakken L, et al. A lexicon for endoscopic adverse events: report of an ASGE workshop. Gastrointest Endosc 2010;71:446-54. 16. Gluck M, Cantone NR, Brandabur JJ, et al. A twenty-year experience with endoscopic therapy for symptomatic primary sclerosing cholangitis. J Clin Gastroenterol 2008;42:1032-9.
Volume
-,
No.
-
: 2014 GASTROINTESTINAL ENDOSCOPY 9
ERCP-related adverse events and PSC
Navaneethan et al
17. Baluyut AR, Sherman S, Lehman GA, et al. Impact of endoscopic therapy on the survival of patients with primary sclerosing cholangitis. Gastrointest Endosc 2001;53:308-12. 18. van den Hazel SJ, Wolfhagen EH, van Buuren HR, et al. Prospective risk assessment of endoscopic retrograde cholangiography in patients with primary sclerosing cholangitis. Dutch PSC Study Group. Endoscopy 2000;32:779-82. 19. James PD, Kaplan GG, Myers RP, et al. Decreasing mortality from acute biliary diseases that require endoscopic retrograde cholangiopancreatography: a nationwide cohort study. Clin Gastroenterol Hepatol. Epub 2013 Oct 2. 20. Dumonceau JM, Andriulli A, Deviere J, et al. European Society of Gastrointestinal Endoscopy (ESGE) Guideline: prophylaxis of postERCP pancreatitis. Endoscopy 2010;42:503-15. 21. Etzel JP, Eng SC, Ko CW, et al. Adverse events after ERCP inpatients with primary sclerosing cholangitis. Gastrointest Endosc 2008;67:643-8. 22. Freeman ML, DiSario JA, Nelson DB, et al. Risk factors for post-ERCP pancreatitis: a prospective, multicenter study. Gastrointest Endosc 2001;54:425-34.
10 GASTROINTESTINAL ENDOSCOPY Volume
-,
No.
-
: 2014
23. Wang P, Li ZS, Liu F, et al. Risk factors for ERCP-related complications: a prospective multicenter study. Am J Gastroenterol 2009;104: 31-40. 24. Alkhatib, Hilden K, Adler DG. Comorbidities, sphincterotomy, and balloon dilation predict post-ERCP adverse events in PSC patients: operator experience is protective. Dig Dis Sci 2011;56: 3685-8. 25. Lella F, Bagnolo F, Colombo E, et al. A simple way of avoiding post-ERCP pancreatitis. Gastrointest Endosc 2004;59:830-4. 26. Cotton PB, Garrow DA, Gallagher J, et al. Risk factors for complications after ERCP: a multivariate analysis of 11,497 procedures over 12 years. Gastrointest Endosc 2009;70:80-8. 27. Testoni PA, Mariani A, Giussani A, et al. Risk factors for post-ERCP pancreatitis in high and low-volume centers and among expert and non-expert operators: a prospective multicenter study. Am J Gastroenterol 2010;105:1753-61. 28. Halttunen J, Keränen I, Udd M, et al. Pancreatic sphincterotomy versus needle knife precut in difficult biliary cannulation. Surg Endosc 2009;23:745-9.
www.giejournal.org
