Endoscopic retrograde cholangiopancreatography (ERCP) is a valuable diagnostic and therapeutic tool in the management of pancreaticobiliary disease. Pancreatitis is a recognized complication of ERCP. Most series (including both pediatric and adult patients) report the prevalence of post-ERCP pancreatitis at 2% to 7%, with some series reporting a prevalence as high as 40% (1–4). This wide range in prevalence can be attributed to the differences in diagnostic criteria used to define post-ERCP pancreatitis as well as the differences in patient populations being studied (eg, increased post-ERCP pancreatitis in patients with sphincter of Oddi dysfunction). In 1991, a consensus statement primarily on the basis of data collected from adult patients was published to define post-ERCP pancreatitis as a rise in serum amylase at least 3 times the upper limit of normal with unplanned or prolonged hospitalization (5,6).
In addition, the severity of post-ERCP pancreatitis was also defined as mild, moderate, or severe. Mild disease requires an unplanned hospitalization or prolonged hospitalization of 2 to 3 days; moderate disease requires an unplanned hospitalization or prolonged hospitalization of 4 to 10 days; and severe disease requires an unplanned hospitalization or prolonged hospitalization for more than 10 days, intervention (either percutaneous or operative), findings of pancreatic necrosis, or pancreatic pseudocyst formation (5,6).
ERCP is performed much less frequently in the pediatric population and the prevalence of post-ERCP pancreatitis in this population is poorly defined using established criteria (5,7). In previously published series of pediatric patients undergoing ERCP, post-ERCP pancreatitis has been reported to be as high as 10% to 29%; however, only 1 of these reports used accepted criteria to define post-ERCP pancreatitis (8–11). In this study, we report the prevalence of post-ERCP pancreatitis in a large number of pediatric patients using accepted criteria.
MATERIALS AND METHODS
After obtaining institutional review board approval, the Mayo Clinic, Rochester, medical and surgical database was queried from 1980 through 2006 to identify all ERCPs performed in patients younger than 18 years. The medical record for each patient was then retrospectively reviewed to identify those procedures that were complicated by post-ERCP pancreatitis as defined by the 1991 consensus statement. Although this established definition was based primarily on adult data, we used pediatric reference ranges for amylase and lipase to determine whether values met the 3 times the upper limit of normal criterion for post-ERCP pancreatitis in addition to an unplanned or prolonged hospitalization. Additional patient and procedural data for ERCPs not complicated by pancreatitis were also collected and reviewed to identify risk factors for post-ERCP pancreatitis.
Patients with chronic pancreatitis were analyzed separately as this subset of patients often has preexisting pancreatitis, ongoing pain issues, and serum amylase levels, which may be difficult to interpret (due to pancreatic fibrosis and insufficiency) (12,13). In this subset of patients, we report the prevalence of recurrent or exacerbated pancreatitis using our own modified criteria defined as pain greater than baseline for more than 24 hours postprocedure necessitating or prolonging hospitalization (14,15).
ERCPs were performed by staff gastroenterologists specialized in interventional and complicated endoscopy with significant experience specifically with ERCP. Adult endoscopes (Olympus TJF 160; Center Valley, PA) are typically used. For children younger than 1 year, smaller endoscopes are used. Patients not hospitalized before ERCP are observed in a postprocedure unit and dismissed home that day after meeting dismissal criteria—tolerating oral intake, no complaints of pain, and alert and oriented. Those patients not meeting discharge criteria are admitted for further evaluation as indicated.
Statistical analysis was performed using the JMP statistical package (JMP Software, Cary, NC). Fisher exact test or χ2 test was used to compare groups for univariate analysis when appropriate on the basis of the group size. Multivariate analysis was conducted using a fit model for those factors with a P ≤ 0.2. A P value <0.05 was considered statistically significant. Values are expressed as mean ± standard deviation.
A total of 343 consecutive pediatric ERCPs were performed in 224 patients at our institution from 1980 through 2006. All ERCPs were performed by adult gastroenterologists with extensive ERCP experience. Two hundred seventy-six ERCPs were performed (in 201 patients) in the absence of chronic pancreatitis, 9 patients developed post-ERCP pain that prompted hospitalization for analgesia, nothing per oral, and intravenous hydration due to a concern for pancreatitis (prevalence 3.3%). Seven of these patients met established criteria for post-ERCP pancreatitis for a prevalence of 2.5%.
Mean age (±SD) of patients developing post-ERCP pancreatitis was 11 ± 6 years (range 2–17 years). Three patients had undergone previous ERCP (each uncomplicated by pancreatitis). Indications included biliary obstruction (n = 5), known choledochal cyst (n = 1), and abdominal pain (n = 1). Six procedures were therapeutic, and 2 patients underwent evaluation of the biliary tree alone (for additional patient characteristics, see Table 1).
All patients developed postprocedural abdominal pain. Two patients had been dismissed home and required readmission; the remaining patients were hospitalized at the time of ERCP. The mean peak serum amylase level was 1006 U/L (range 337–3197 U/L; normal reference range 21–110 U/L), and the mean lipase level was 1267 U/L (range 88–2694 U/L; normal reference range 10–73 U/L). All of the patients were managed conservatively with bowel rest and intravenous fluids. No patient required further intervention for pancreatitis (including the 2 patients who did not meet criteria for post-ERCP pancreatitis). The mean duration of hospitalization was 4 days (range 2–7 days). According to the defined severity of post-ERCP pancreatitis, there were no episodes of severe pancreatitis; 5 patients developed moderate pancreatitis and 2 developed mild pancreatitis.
Pancreatic duct stents were placed in 32 patients undergoing uncomplicated ERCP; 1 patient in the post-ERCP pancreatitis group underwent pancreatic duct stenting. Patients undergoing balloon dilation were more likely to develop post-ERCP pancreatitis by univariate analysis (P = 0.02). The indications for balloon manipulation in these 3 patients included biliary stones (n = 2) or stricture (n = 1). Patients who underwent diagnostic-only ERCP were less likely to develop post-ERCP pancreatitis (P = 0.01); this was the only significant predictor of post-ERCP pancreatitis after multivariate analysis as well (P < 0.01; see Table 2).
Chronic Pancreatitis Patients
Sixty-seven procedures were performed in 23 children diagnosed with chronic pancreatitis, which was confirmed radiologically by ERCP, endoscopic ultrasound, and/or computed tomography. Ten of these patients developed post-ERCP pain that was worse from their baseline for more than 24 hours requiring parenteral narcotics for pain control and hospitalization for a prevalence of 14.9% (for additional patient data, see Table 3). Two of these patients had been dismissed home and required readmission, and mean duration of hospitalization for all of the patients was 4 days (range 2–8 days). The remaining 8 patients had either been hospitalized at the time of ERCP or were admitted in anticipation of postprocedural pain. Stent placement was the only predictive factor (by univariate and multivariate analyses) for post-ERCP pain in this patient group (P = 0.02, see Table 4). Patients with post-ERCP pain were also more likely to have undergone a previous ERCP (60% vs 21%, P = 0.005). The post-ERCP pain was self-limited in all of these patients.
Post-ERCP pancreatitis in the pediatric population is not as well defined as in the adult population, and the existing literature suggests a relatively high rate of post-ERCP pancreatitis (8–11). In our review of 343 ERCPs performed in pediatric patients, post-ERCP pancreatitis, defined by using established criteria, occurred in less than 3%. Furthermore, the low prevalence in our series mirrors that reported in 1 other large series of pediatric ERCP where the prevalence of post-ERCP pancreatitis was 3.4% (7). This is far lower than what has been reported in smaller series and demonstrates that ERCP is safe in children when performed by experienced endoscopists. However, it must be acknowledged that our experience may differ from other reports because the ERCPs were performed by adult gastroenterologists, and we used a standard definition of post-ERCP pancreatitis. Furthermore, this analysis is limited by its retrospective nature. Nonetheless, this is a large series of pediatric ERCP that not only supports the safety of ERCP in children but also identifies risk factors for post-ERCP pancreatitis.
Using univariate analysis, we identified several risk factors for the development of pancreatitis—balloon manipulation (either dilation or stone retrieval) and therapeutic ERCP. On multivariate analysis, therapeutic ERCP was the only factor that remained a significant predictor of post-ERCP pancreatitis. With the improvements in radiologic imaging and rapid decline in the frequency with which diagnostic ERCP is performed, it is possible that the growing volume of therapeutic ERCPs will lead to a higher prevalence of post-ERCP pancreatitis. In fact, of all of the patients with post-ERCP pancreatitis only 2 occurred before 2003. From 1980 through 2002, the incidence was 1%; from 2003 through 2006, the incidence was 6%. The rate of therapeutic ERCP in those time periods also increased from 30% to 77%.
It is clear from these data that manipulating the biliary or pancreatic ducts increases the likelihood of developing post-ERCP pancreatitis, but the exact pathophysiological mechanisms of how this induces pancreatitis are not clear. Although once felt to be related to pressure within the pancreatic duct, the lower rate of pancreatitis in diagnostic-only procedures as well as several cases of post-ERCP pancreatitis after therapeutic procedures where the pancreatic duct was not manipulated or evaluated with contrast injection are compelling. This could imply that post-ERCP pancreatitis is related to trauma and/or edema at the papillary orifice. This is the reasoning behind the current practice of placing prophylactic pancreatic stents in patients at increased risk for post-ERCP pancreatitis (16–20). However, we did not find a protective benefit for pancreatic duct stents in this series, although the numbers within this subset are small and the pancreatic stents were not placed specifically for prophylaxis of post-ERCP pancreatitis, but rather as a therapy for pancreatic ductal disease. Additionally, in the chronic pancreatitis group, pancreatic duct stent placement increased the likelihood of developing post-ERCP pancreatitis.
The use of prophylactic nonsteroidal anti-inflammatory drugs has also been found to prevent post-ERCP pancreatitis in selected patients (21,22). This further supports the theory of edema, or another inflammatory process, incited by manipulation of the biliopancreatic ductal system mediating post-ERCP pancreatitis.
Fortunately, all of the cases of post-ERCP pancreatitis in our population were self-limiting. Using an established definition for the severity of pancreatitis, none of the patients with post-ERCP pancreatitis developed severe disease. The majority were classified as moderate severity (pancreatitis with hospitalization of 4–10 days) and the remainder had mild severity (pancreatitis with hospitalization of 2–3 days).
Patients with chronic pancreatitis pose a significant challenge in defining post-ERCP pancreatitis using the established definition because serum amylase and lipase levels may not be reflective of acute pancreatic injury when there is underlying fibrosis of the pancreas with subsequent exocrine insufficiency. Therefore, we applied a modified definition for post-ERCP pancreatitis in this subset of patients using post-ERCP pain that was worse than baseline for more than 24 hours postprocedure requiring either an unplanned hospitalization or prolonged hospitalization (14,15). On the basis of this modified criteria, these patients experience a much higher rate of post-ERCP pancreatitis of 14.9%. As mentioned previously, pancreatic duct stenting was a risk factor for chronic pancreatitis patients to develop post-ERCP pain. Additionally, having had a prior ERCP (complicated or uncomplicated) seemed to predispose patients with chronic pancreatitis to post-ERCP pain. The reason for this is unclear. It may simply relate to underlying disease progression or reflect the high prevalence of post-ERCP pancreatitis in this patient population.
It has been our more recent practice to not routinely admit patients after ERCP unless they are at high risk for post-ERCP complications, specifically pancreatitis, on the basis of a prior history of post-ERCP pancreatitis, complexity of the procedure, and/or known chronic pancreatitis. Overall, only 4 patients (of 17) required readmission for post-ERCP pancreatitis. The remaining 13 patients had either been identified as high risk or ERCP was part of an inpatient workup. We feel that our low rate of post-ERCP pancreatitis justifies performing ERCP on an outpatient basis.
In conclusion, the prevalence of post-ERCP pancreatitis in pediatric patients is far lower than previously reported and there should be little concern in referring pediatric patients for ERCP when indicated. However, therapeutic ERCP does carry a higher, but still acceptable, risk of post-ERCP pancreatitis. In patients with chronic pancreatitis, the prevalence of post-ERCP pain requiring hospitalization is higher yet, although routine admission is still probably not warranted in these patients.
1. Cooper ST, Slivka A. Incidence, risk factors, and prevention of post-ERCP pancreatitis. Gastroenterol Clin North Am 2007; 36:259–276.
2. Andriulli A, Loperfeld S, Napolitano G, et al. Incidence rates of post-ERCP complications: a systematic summary of prospective studies. Am J Gastroenterol 2007; 102:1781–1788.
3. Redy N, Wilcox CM, Tamhane A, et al. Protocol-based medical management of post-ERCP pancreatitis. J Gastroenterol Hepatol 2008; 23:344–347.
4. Cooper ST, Slivka A. Incidence, risk factors, and prevention of post-ERCP pancreatitis. Gastroenterol Clin North Am 2007; 36:259–276.
5. Cotton PB, Lehman G, Vennes J, et al. Endoscopic sphincterotomy complications and their management: an attempt at consensus. Gastrointest Endosc 1991; 37:383–393.
6. Freeman ML, Guda NM. Prevention of post-ERCP pancreatitis: a comprehensive review. Gastrointest Endosc 2004; 59:845–864.
7. Varadarajulu S, Wilcox CM, Hawes RH, et al. Technical outcomes and complications of ERCP in children. Gastrointest Endosc 2004; 60:367–371.
8. Cheng CL, Fogel EL, Sherman S, et al. Diagnostic and therapeutic endoscopic retrograde cholangiopancreatography in children: a large series report. Pediatr Gastroenterol Nutr 2005; 41:445–453.
9. Johanson JF, Cooper G, Eisen GM, et al. Quality assessment of ERCP. Endoscopic retrograde cholangiopacreatography. Gastrointest Endosc 2002; 56:165–169.
10. Guelrud M, Mujica C, Jaen D, et al. The role of ERCP in the diagnosis and treatment of idiopathic and recurrent pancreatitis in children and adolescents. Gastrointest Endosc 1994; 40:428–436.
11. Prasil P, Laberge JM, Barkun A, et al. Endoscopic retrograde cholangiopancreatography in children: a surgeon's perspective. J Pediatr Surg 2001; 36:733–735.
12. Etemad B, Whitcomb DC. Chronic pancreatitis: diagnosis, classification, and new genetic developments. Gastroenterol 2001; 120:682–707.
13. Ammann RW, Heitz PU, Kloppel G. Course of alcoholic chronic pancreatitis: a prospective clinicomorphological long-term study. Gastroenterology 1996; 111:224–231.
14. Rabenstein T, Eckhart GH. Post-ERCP pancreatitis: is the endoscopist's experience the major risk factor? J Pancreas 2002; 3:177–187.
15. Rabenstein T, Schneider HT, Hahn EG, et al. 25 years of endoscopic sphincterotomy in Erlangen: assessment of experience in the treatment of 3498 patients. Gastroenterology 1998; 30:A194–A201.
16. Simmons DT, Peterson BT, Baron TH, et al. Risk of pancreatitis following endoscopically place large-bore plastic biliary stents with and without biliary sphincterotomy for management of postoperative bile leaks. Surg Endosc 2008; 22:1459–1463.
17. Saad M, Fogel EL, McHenry L, et al. Pancreatic duct stent placement prevents post-ERCP pancreatitis in patients with suspected sphincter of Oddi dysfunction but normal manometry results. Gastrointest Endosc 2008; 67:255–261.
18. Freeman ML. Pancreatic stents for prevention of post-endoscopic retrograde cholangiopancreatography pancreatitis. Clin Gastroenterol Hepatol 2007; 5:1354–1365.
19. Tsuchiya T, Itoi T, Sofuni A, et al. Temporary pancreatic stent to prevent post endoscopic retrograde cholangiopancreatography pancreatitis: a preliminary, single-center, randomized controlled trial. J Hepatobiliary Pancreat Surg 2007; 14:302–307.
20. Varadarajulu S, Wilcox CM. Endoscopic management of sphincter of Oddi dysfunction in children. J Pediatr Gastroenterol Nutr 2006; 42:526–530.
21. Cheon YK, Cho KB, Watkins JL, et al. Efficacy of diclofenac in the prevention of post-ERCP pancreatitis in predominantly high-risk patients: a randomized double-blind prospective trial. Gastrointest Endosc 2007; 66:1126–1132.
22. Elmunzer BJ, Waljee AK, Elta GH, et al. A meta-analysis of NSAIDs in the prevention of post-ERCP pancreatitis. Gut 2008; 57:1262–1267.
© 2009 Lippincott Williams & Wilkins, Inc.