Endoscopic Retrograde Cholangiopancreatography in Children: A Surgeons Perspective

By Pascale Prasil, Jean-Martin Laberge, Alan Barkun, and He le ` ne Flageole Montreal, Quebec

Purpose: The aim of this study was to review the indications, success rate, and complications of endoscopic retrograde cholangiopancreatography (ERCP) in the pediatric age group. Methods: From 1990 to 1999, 21 ERCP procedures were attempted in 20 patients. They consisted of 8 boys and 12 girls whose age ranged from 4 to 17 years (mean, 11.3 years). Fourteen were performed under deep sedation (mean age, 12.8 years), and 7 were done under general anesthesia (mean age, 7.6 years). All ERCP procedures were performed by experienced adult endoscopists. Results: The indication for ERCP was biliary in 15 patients. Eleven had suspected choledocholithiasis by either ultrasound scan, intraoperative cholangiogram or magnetic resonance imaging (MRI). In 6 cases, the ERCP was done for pancreatic pathology. In 11 patients, the ERCP was diagnostic only, and in 10 a therapeutic procedure was done. The

overall success rate was 90.5%. Post-ERCP complications consisted of 6 episodes of pancreatitis (28.5%), 4 of which followed a therapeutic procedure, and 1 episode of bleeding. Pancreatitis resolved 2 to 6 days post-ERCP. The patients underwent follow-up between 2 and 56 months after the ERCP (mean, 11 months).

Conclusions: The authors conclude that even in experienced hands, ERCP in the pediatric population has a much higher complication rate than in adults (33.3%). We recommend that very specic indications be met before subjecting a pediatric patient to an endoscopic retrograde cholangiopancreatography. J Pediatr Surg 36:733-735. Copyright 2001 by W.B. Saunders Company.
INDEX WORDS: Endoscopic retrograde cholangiopancreatography, complications.

NDOSCOPIC retrograde cholangiopancreatography (ERCP) has been used increasingly in the pediatric population over the last 10 years both as a diagnostic and therapeutic modality. Several small and very few larger series have been published on ERCP in children. They report success and complication rates comparable with those quoted in the adult literature.1-7 After encountering several cases of pancreatitis after ERCP, some that were done for suspected rather then proven common bile duct (CBD) stones, we reviewed the experience with ERCP at the Montreal Childrens Hospital (MCH) to examine our indications, success rate, and complications in comparison with those of other series, and to determine whether our indications should be revised.
MATERIALS AND METHODS
We reviewed retrospectively the charts of all patients who underwent an ERCP from 1990 to 1999. Twenty-one ERCP procedures were undertaken in 20 patients, comprising 8 boys and 12 girls whose ages ranged from 4 to 17 years (mean, 11.3 years). The indications, results, and complications were noted as well as the type of anesthesia used. The latter was decided according to the age of the patient and expected cooperation. The 7 youngest patients (mean age, 7.6 years) were treated under general anesthesia in the interventional radiology suite with the assistance of a pediatric anesthetist, whereas the 14 older teenagers (mean age, 12.8 years) were given sedation. All ERCP procedures, whether diagnostic or therapeutic, were performed by a very experienced adult endoscopist using an adult gastroduodenoscope. In the complications, pancreatitis was dened as abdominal pain post-ERCP
Journal of Pediatric Surgery, Vol 36, No 5 (May), 2001: pp 733-735

associated with any elevation of the pancreatic enzymes (amylase and lipase).

RESULTS

The two main indications for ERCP were biliary pathology in 15 cases and pancreatic pathology in 6 cases. In the biliary group, 11 patients had suspected or proven common bile duct (CBD) stones, and 3 patients had unexplained biliary tract dilatation by either ultrasound scan, magnetic resonance imaging (MRI), or intraoperative cholangiogram. The 15th patient had the examination done to rule out sclerosing cholangitis. The ndings at ERCP in this biliary pathology group were 9 common bile duct stones, 3 normal examinations, 1 CBD stricture, 1 choledochal cyst, and 1 patient in whom the CBD could not be cannulated. For the 9 patients with
From the Division of General Pediatric Surgery, Montreal Childrens Hospital, McGill University Health Center, Montreal, Quebec, Canada. Presented at the 32nd Annual Meeting of the Canadian Association of Paediatric Surgeons, Cha teau Montebello, Quebec, Canada, September 15-18, 2000. Address reprint requests to He le ` ne Flageole, MD, FRCS(C), FACS, Montreal Childrens Hospital, 2300 Tupper St, Room C-1129, Montreal, Quebec, Canada H3H 1P3. Copyright 2001 by W.B. Saunders Company 0022-3468/01/3605-0014$35.00/0 doi:10.1053/jpsu.2001.22948
733

734

PRASIL ET AL

CBD stones, 8 sphincterotomies with stone extraction were done, and 1 underwent partial papillotomy without stone extraction. The patient with the CBD stricture had a sphincterotomy with placement of a stent in the CBD. Overall, 5 ERCP procedures were diagnostic and 10 therapeutic in this group with an overall success rate (dened as ability to complete the diagnostic or therapeutic procedure) of 86.7%. Seven patients in this group suffered complications post-ERCP (47%). Six of them had pancreatitis as evidenced by abdominal pain and biochemical abnormalities, namely elevated serum amylase and lipase levels. Of these, 4 children had undergone a sphincterotomy. One other patient had a simple diagnostic ERCP, the results of which were normal, and in the sixth patient there was inability to cannulate the CBD. Two of the episodes of pancreatitis were severe enough to require hospital stays of 8 and 9 days, whereas the other 4 episodes were rather mild with complete resolution of the symptoms in 1 to 3 days. The other complication in the biliary pathology group consisted of 1 episode of bleeding. The patient returned to hospital 24 hours after ERCP with melena and hypotension. He required transfusion of 2 units of packed red blood cells, and the bleeding ceased spontaneously. There were 6 ERCP procedures performed for pancreatic pathology in 5 patients with either recurrent or chronic pancreatitis. One child underwent 2 ERCP procedures several years apart. All procedures were diagnostic only, and the success rate in this group was 100%. These showed 3 normal ndings and 3 pancreatic duct anomalies, one of which was amenable to surgical therapy. There were no complications in this group. The follow-up after ERCP ranged from 2 to 56 months (mean, 11 months). The results are summarized in Table 1. None of the patients had long-term sequelae from ERCP, regardless of whether they had early complications.
DISCUSSION

ERCP is being used with increasing frequency in the pediatric and even the neonatal population. In biliary disorders, it can be both diagnostic and therapeutic, especially in cases of choledocholithiasis. However, in cases in which the need for therapeutic intervention is uncertain, magnetic resonance cholangiopancreatograTable 1. Summary of Results
Indication Biliary Pathology (n 15) Pancreatic Pathology (n 6)

Diagnostic ERCP Therapeutic ERCP Success rate Pancreatitis post-ERCP Bleeding post-ERCP

5 10 86.7% 6 (40%) 1 (6.6%)

6 0 100% 0% 0%

phy (MRCP) is being used increasingly as a diagnostic modality prior to ERCP.8,9 ERCP also is very useful in dening the anatomy of pancreatic duct abnormalities, where the results could dictate the therapeutic options offered to the patient. Examples would include the multiple strictures and dilatations sometimes seen in chronic pancreatitis and pancreas divisum causing recurrent pancreatitis, conditions amenable to surgical correction. Nonetheless, ERCP is not without risks, especially in young children and infants. As do others, we feel that in this group of patients, a general anesthetic is the safest method to protect the airway and ensure an immobile patient.6,7,10 In our series, 35% (7 of 20) of patients belonged to this category. There are reports of ERCP being performed in the neonate and young child under sedation, but this approach has yet to gain wide acceptance.1 In such a context, the indication for ERCP should be strong. It should not be used as a screening test for conditions such as unexplained abdominal pain, in which the likelihood of nding signicant pathology is minimal.3 When examining complications from ERCP, the reference point clearly is the adult literature. Even in adults, the morbidity and mortality rates after ERCP are appreciable. A prospective multicenter study by Loperdo et al11 conducted on 2,769 consecutive patients in 9 different centers makes that point. They reported major complications in 4% of patients, with pancreatitis, cholangitis, and hemorrhage being the most frequent. There were 1.38% major complications and 0.21% deaths in the diagnostic ERCP group, whereas patients in the therapeutic group suffered 5.4% complications and 0.49% mortality rate. Centers performing fewer than 200 ERCP per year and the performance of a partial papillotomy, sometimes referred to as a pre-cut procedure, were identied as independent risk factors for complications. In the pediatric population, the relatively low volume of cases denitely is an issue. This problem is minimized in our institution by having a very limited number of very experienced adult endoscopists perform the procedures in children.12 These selected individuals perform a large number of adult ERCP procedures each year, but their level of comfort with children undoubtedly varies because we only have a few cases each year. Even in these experienced hands, our complication rate was 33.3%, 86% of which were episodes of pancreatitis. One third of these episodes were severe enough to require hospitalization for more than 1 week. Half of patients undergoing a therapeutic ERCP procedure suffered from complications. Our rate of complications seems higher than that of most reported series despite all the measures taken to minimize the risks as described above. This could be partially explained by the fact that we were very rigorous in reporting them. Each patient had routine biochemical testing the day after ERCP, and any eleva-

ERCP: A SURGEONS PERSPECTIVE

735

tion in amylase and lipase levels were counted as pancreatitis. Four of our 6 patients had mild pancreatitis, which could perhaps have gone unnoticed in other circumstances. Nonetheless, the complication rate from ERCP is signicantly higher in our population than that reported in adults. We therefore recommend that strict selection criteria be met before subjecting a pediatric patient to an ERCP, especially when biliary lithiasis is the indication. Suspicion of CBD stones on the basis of CBD dilatation

or elevated enzymes should not be an indication for preoperative ERCP. In such cases, we recommend proceeding rst with laparoscopic cholecystectomy and intraoperative cholangiogram. ERCP follows if CBD stones persist despite intraoperative ushing. ERCP remains an essential diagnostic tool in children with unexplained recurrent pancreatitis and in those with biliary tract pathology such as Carolis disease, sclerosing cholangitis, and choledochal cyst when less invasive imaging modalities fail to provide a diagnosis.

REFERENCES
1. Guelrud M: Endoscopic retrograde cholangiopancreatography in children. The Gastroenterologist 4:81-97, 1996 2. Guelrud M, Mendoz S, Jaen D, et al: ERCP and endoscopic sphincterotomy in infants and children with jaundice due to common bile duct stones. Gastrointest Endosc 38:450-453, 1992 3. Brown CW, Werlin SL, Geenen JE, et al: The diagnostic and therapeutic role of endoscopic retrograde cholangiopancreatography in children. J Pediatr Gastroenterol Nutr 17:19-23, 1993 4. Cotton PB, Laage NJ: Endoscopic retrograde cholangiopancreatography in children. Arch Dis Child 57:131-136, 1982 5. Buckley A, Connon JJ: The role of ERCP in children and adolescents. Gastrointest Endosc 36:369-372, 1990 6. Putnam PE, Kocoshis SA, Orenstein SR, et al: Pediatric endoscopic retrograde cholangiopancreatography. Am J Gastroenterol 86: 824-830, 1991 7. Brown KO, Goldschmiedt M: Endoscopic therapy of biliary and pancreatic disorders in children. Endoscopy 26:719-723, 1994 8. Mehta SN, Reinhold C, Barkun AN: Magnetic resonance cholangiopancratography. GI Clin North Am 7:247-270, 1997 9. Tekehara Y: Can MRCP replace ERCP? J Mag Res Imaging 8:517-534, 1998 10. Teng RF, Yokohata K, Utsunomiya N, et al: Endoscopic retrograde cholangiopancreatography in infants and children. J Gastroenterol 35:39-42, 2000 11. Loperdo S, Angelini G, Benedetti G, et al: Major early complications from diagnostic and therapeutic ERCP: A prospective multicenter study. Gastrointest Endosc 48:1-10, 1998 12. Mehta SN, Pavone E, Barkun AN: Out-patient therapeutic ERCP: A series of 262 consecutive cases. Gastrointest Endosc 44:443449, 1996