Endoscopic retrograde cholangiopancreatography in children. Subcommittee on Endoscopy and Procedures of the Patient Care Committee of the North American Society for Pediatric Gastroenterology and Nutrition.

Abstract

Endoscopic retrograde cholangiopancreatography (ERCP) is a specialized procedure that combines the skills of gastrointestinal endoscopy and interpretive fluoroscopy for diagnostic and therapeutic management of biliary tract and pancreatic disorders. The procedure has been widely applied in adult patients for more than 20 years and has recently become an accepted modality in pediatrics. Pediatric centers that offer ERCP encounter several obstacles, including few pediatric subspecialists adequately trained to perform the procedure independently, insufficient numbers of patient to maintain the necessary endoscopic skills for diagnostic and therapeutic ERCPs, and high equipment cost. A clear consensus has not yet emerged among pediatricians and pediatric gastroenterologists for appropriate indications or practice guidelines for this procedure. The Subcommittee on Endoscopy and Procedures of the North American Society for Pediatric Gastroenterology and Nutrition (NASPGN) has prepared the following review and position statement regarding ERCP in children. This document is expected to undergo periodic revision to reflect future clinical experience and changing technology. BACKGROUND In 1968, McCune et al. (1) reported the first successful cannulation of the ampulla of Vater using a gastroscope. Adult size (10-mm diameter) side-viewing duodenoscopes became available in the 1970s. In 1976, Waye (2) reported the first successful ERCP in a child, using the adult-sized duodenoscope in a 3.5-month-old, 6.0-kg infant with cholestatic liver disease. Several additional pediatric cases (3–5) and one small (n = 18) pediatric series (6) were reported using this endoscope. Smaller diameter duodenoscopes, 7.5 to 9.0 mm, were developed in the 1980s and 1990s. Refinements in design and expanded distribution of these duodenoscopes have fostered application of ERCP in children at major pediatric referral centers throughout the world. Published experience, however, remains relatively limited. Although feasibility of ERCP in young children and infants has been repeatedly demonstrated in case reports and small series, safety and clinical outcomes for diagnostic and therapeutic interventions have not been critically examined in large numbers of pediatric patients. Even the largest U. S. series by Brown et al. in 1993 (7) included fewer than 100 patients. Table 1 summarizes details including age range, indications, and rates of complications for the larger published pediatric series (6–18,29).TABLE 1: Pediatric ERCP seriesTECHNICAL CONSIDERATIONS ERCP is a technically demanding procedure that has developed significantly during the past 20 years in advanced diagnostic and therapeutic capabilities. Practitioners should have a full appreciation of the technical complexity and limitations of this procedure. Patient Preparation and Sedation Children undergoing ERCP require preparation and sedation similar to that used in other endoscopic procedures, including explanation of the procedure to the child in an age-appropriate manner, and informed consent from the guardian. The depth and duration of sedation may exceed the level required for diagnostic upper gastrointestinal endoscopy, depending on such factors as the skill of the endoscopist, the anatomic findings, and the technical difficulty of the intervention. The relatively large diameter of standard side-viewing endoscopes (e.g., 11–12 mm) may cause discomfort and compress the soft-walled trachea in young children. Although using a smaller diameter (7.5 mm) duodenoscope may minimize this problem, some accessories are not compatible with this size instrument. Prone positioning during ERCP also compromises chest and lung excursion and may result in hypoventilation and hypoxia in a sedated child. Given these considerations, general anesthesia with endotracheal intubation may be appropriate for some children undergoing ERCP. Insufficient outcome data exist to make firm recommendations about the comparative adequacy, safety, and cost of intravenous sedation versus general anesthesia for children who undergo ERCP. Adequately trained personnel for patient monitoring are required with either method of sedation. Fluoroscopy Fluoroscopic equipment should be suitable for children of all ages with settings optimized to produce real-time and hard copy images while minimizing radiation exposure. Modern digital equipment is ideal but may not always be available and is not essential (19). Portable radiograph may not result in ideal resolution for subtle lesions but may be required for intraoperative studies. The endoscopist should have basic knowledge of radiation safety and take necessary precautions to protect patients and personnel from unnecessary radiation exposure. If possible, a licensed radiation technologist with pediatric experience should assist with fluoroscopy, and a pediatric radiologist should be available for consultation in situations of unusual anatomy. A variety of water-soluble low-osmolar (nonionic) and high-osmolar contrast media are commercially available and may be used for ERCP in children as in adults (20). An optimal iodine concentration has not been defined, but a range of 150 to 300 mg iodine per milliliter is commonly used. Low-osmolar contrast media provide no added protection against ERCP-related pancreatitis compared with high-osmolar media, but they may lower the risk of systemic contrast reaction in patients with a history of anaphylactic reaction (20). Procedure Setting Diagnostic ERCP is associated with a low risk of complications (discussed later). Therefore, selected cases may be appropriate for a non–hospital-based ambulatory medical center. Overnight observation in the hospital may be indicated after therapeutic ERCP, given the greater potential risk for serious complications. Access to subspecialists in the fields of pediatric surgery, anesthesia, and radiology is necessary. Nursing personnel should be experienced or educated to recognize the signs of endoscopic complications in children and to expedite supportive intervention. Because the onset of post-ERCP pancreatitis may be delayed by up to 12 hours after the procedure (21), patients should remain close to a hospital during this time. Endoscopic Equipment Currently available duodenoscopes permit diagnostic ERCP and, at least, limited therapeutic ERCP in children of virtually any age, including neonates. ERCP in infants and children less than 2 years of age is optimally performed using a specially designed, small-diameter duodenoscope (9,22). The latest version of this scope has a cannula elevator and a 2.0-mm diameter operating channel that accepts cannulas or other accessories that are 5 Fr or smaller. Sphincterotomy and stone extraction can be performed using this small endoscope and commercially available accessories. A standard diagnostic or medium-diameter (11.0–12.0 mm) duodenoscope with a 3.2-mm operating channel may be used in most children more than 2 to 3 years of age. The channel accommodates medium-sized stents and passage of dilating catheters up to 7 Fr. Therapeutic large-diameter (>13 mm) endoscopes with a 4.2-mm channel are infrequently required to place 10 Fr stents in larger adolescents. Endoscopy Assistants An endoscopist performing ERCP should have at least one assistant with knowledge of the specialized accessories and techniques used during ERCP. The assistant may have either nursing or technical training but should not be relied on to provide patient support during the procedure. If intravenous sedation is used, additional nursing support is required for patient monitoring, as recommended by the American Academy of Pediatrics (23). A nurse skilled in airway management should remain at the head of the patient, supporting the airway, monitoring vital signs, and administering additional medications as needed while a second assistant assists with the endoscopic procedure. Role of the Specially Trained Endoscopist ERCP should be not be undertaken in any age patient by an endoscopist who has insufficient training and experience to successfully complete the procedure in 80% to 90% of cases. Guidelines issued in 1992 by the American Society of Gastrointestinal Endoscopy (ASGE) (24) recommended performance of 100 ERCPs (75 diagnostic and 25 therapeutic) as a minimum threshold for assessing competence during training. Recently published studies have challenged the adequacy of this threshold number indicating that more than 200 procedures are required to reliably achieve deep common bile duct cannulation with 80% probability (25,26). Deep common bile duct cannulation is an important prerequisite for many therapeutic interventions. The frequency or volume of ERCP experience that is necessary to maintain competency is unknown. A multicenter prospective study by Freeman et al. (27) demonstrated a significantly higher rate of complications after endoscopic sphincterotomy by endoscopists who perform less than one sphincterotomy per week. The relatively small volume of pediatric ERCP cases in a given center may not provide sufficient experience for a pediatric endoscopist to reach minimum training thresholds or to sustain a level of competence needed for therapeutic ERCP. Therefore, within a medical center a partnership should be formed between pediatric and adult therapeutic endoscopists combining the necessary cognitive and technical skills to provide optimal care for pediatric patients. RISKS AND COMPLICATIONS The subject of ERCP-related complications in adults has been extensively reviewed recently (28). The major risks and complications of ERCP in children are the same as in adults and include pancreatitis, hemorrhage, infection, and perforation. The relative risk for these complications in children is not well established because of the small numbers of patients in pediatric series. One pediatric death caused by infection of a pancreatic pseudocyst after ERCP was reported more than 20 years ago (6). Pancreatitis is the most common complication after ERCP. Among larger pediatric series, pancreatitis has occurred in up to 8% of patients (Table 1). This compares favorably with rates of 3% to 7% reported in adults (28). In a large pediatric ERCP series (7) (121 procedures) the overall complication rate was 11.6%. Pancreatitis occurred in four cases (3.3%). Patients undergoing therapeutic intervention may be at increased risk for post-ERCP pancreatitis, occurring in 17% of patients in one pediatric series (15). Although the pathogenesis of this complication is not well understood, efforts should be made to minimize repeated duct cannulation and injection, prolonged manometry catheter infusion into the pancreatic duct, or other unnecessary mechanical irritation of the papilla or pancreatic duct orifice. Hemorrhage and perforation have rarely been reported after ERCP in children (10,29,30). Rates are likely to be similar to those reported for adults: 0.7% to 2.0% for hemorrhage and 0.3% to 0.6% for perforation (27,28). The rate of ERCP-related bacteremia in children is not known, but in adult patients it is 15% for diagnostic and 27% for therapeutic studies (28). Few cases of clinical infection have been reported in children (7). Guidelines for antibiotic prophylaxis during ERCP have been updated recently (31,32). Antibiotic prophylaxis is recommended in patients with high-risk cardiovascular conditions, including prosthetic valves, a history of endocarditis, systemic-pulmonary shunt, and synthetic vascular grafts less than 1 year old and in any patient with an obstructed bile duct or pancreatic pseudocyst. The incidence of complications associated with therapeutic interventions, such as sphincterotomy or stent placement, is not known for children. A recent multicenter prospective study of complications in adults undergoing endoscopic sphincterotomy showed an overall complication rate of 9.8%(27). The most frequent complications included: pancreatitis (5.4%), hemorrhage (2.0%), cholangitis (1.0%), cholecystitis (0.5%), and perforation (0.3%). Failed or inadequate biliary drainage is a major risk factor for cholangitis. Stent occlusion occurs in 20% to 30% of patients within 3 months of insertion, and the rate of stent migration for both biliary duct and pancreatic duct stents is approximately 5.0%(28). The long-term complications of sphincterotomy or use of stents in pediatric patients are unknown. Contraindications to ERCP include unstable cardiovascular, pulmonary, or neurologic conditions or suspected bowel perforation. Esophageal stricture is a relative contraindication, depending on the severity of the stricture. Coagulopathy is a relative contraindication and should be corrected before ERCP if possible. Medications that may affect platelet function, such as aspirin or nonsteroidal anti-inflammatory drugs (NSAIDs), should be avoided or withheld before the procedure. DIAGNOSTIC AND THERAPEUTIC INDICATIONS Although the indications for ERCP are similar for children and adults, the relative frequency of each indication differs. For example, children are much more likely to have indications related to congenital abnormalities or trauma than to malignancy. Indications related to biliary lithiasis or to prior surgeries are less common in children. Magnetic resonance imaging cholangiopancreatography (MRCP) can provide high-quality noninvasive imaging of biliary and pancreatic ducts in adults and may eliminate the need for diagnostic ERCP in some settings. Initial reports of MRCP in children (33–36) are very promising, but further study is needed to compare accuracy and utility with ERCP for specific disorders. ERCP provides high-resolution imaging of duct anatomy and permits sampling of tissue and fluid as well as therapeutic intervention. Most endoscopic therapies used in adult patients have been performed in children; however, experience with therapeutic ERCP in children remains limited. Depending on available expertise, conventional surgical intervention may be preferred rather than endoscopic intervention for children who are otherwise good surgical candidates. Biliary Indications Biliary indications for ERCP in children are listed in Table 2. Endoscopic investigation of neonatal cholestasis is unique to pediatrics, and the role of ERCP remains controversial. Structural causes of neonatal cholestasis including biliary atresia, choledochal cyst, choledocholithiasis and inspissated bile, intrahepatic bile duct paucity or hypoplasia, neonatal sclerosing cholangitis, and congenital bile duct stricture should be ruled out to minimize ongoing liver injury. Diagnostic tests such as abdominal ultrasound or computed tomography (CT), liver biopsy, biliary scintigraphy, and a biochemical profile cannot always distinguish between intrahepatic cholestasis because of neonatal hepatitis and biliary atresia. Ultimately, a cholangiogram or surgical exploration may be required. ERCP offers a relatively noninvasive alternative to percutaneous transhepatic or transcholecystic cholangiography or open surgical cholangiography, and, if normal biliary anatomy is demonstrated, may eliminate the need for a surgical procedure. Several investigators have reported a high positive predictive value of ERCP in this setting (11,16–18,30). Endoscopic findings (11) that suggest the diagnosis of biliary atresia include: complete absence of contrast in the bile duct despite filling of the pancreatic duct, partial filling of the bile duct with abnormal termination, and no appearance of bile in the duodenum. Guelrud et al. (37) have reported the largest series of infants evaluated by ERCP for neonatal cholestasis. One hundred fifty-five infants were examined during 12 years with successful cannulation in 147 (95%). After observations made during ERCP, the diagnosis was incorrect in only two (1.3%) patients. Together with clinical, biochemical, sonographic, and liver biopsy findings, ERCP may have value under selected circumstances to clarify the cause of neonatal cholestasis.TABLE 2: Biliary indications for ERCPSyndromic (Alagille's syndrome) and nonsyndromic paucity of intrahepatic bile ducts may mimic biliary atresia in the neonatal period; ERCP may be helpful to exclude biliary atresia, because characteristic histopathology of bile duct paucity may not fully develop in infancy (38). The diagnosis of congenital hepatic fibrosis is definitively established by liver biopsy. Although ultrasound or other radiologic imaging can often establish the diagnosis of a choledochal cyst (39), ERCP confirms the diagnosis and assists surgical planning (41). The cholangiographic finding of a long common channel or anomalous pancreatobiliary junction has been frequently associated with and may contribute to the formation of choledochal cyst (41). Choledocholithiasis with bile duct obstruction may mimic the findings of choledochal cyst in young infants (42) and should be excluded by cholangiography. ERCP may be indicated in children with suspected sclerosing cholangitis or cholangiopathy in association with inflammatory bowel disease, histiocytosis X, human immunodeficiency virus infection (43), or other immunodeficiency states. ERCP is also indicated to evaluate strictures arising from operative injury, abdominal trauma, and extrinsic compression from chronic fibrosing pancreatitis or from tumors, either benign or malignant. ERCP has been used to perform sphincter of Oddi manometry in adults and children with suspected biliary dyskinesia or sphincter of Oddi dysmotility. However, given that there are no normal control data in children and minimal published experience and inherent technical difficulties with the test, no recommendations can be made regarding the clinical application or reliability of this test in children. Endoscopic sphincterotomy and stone extraction for choledocholithiasis is the most commonly used biliary intervention in children. This has been performed successfully in infants as young as 2 months (44–46), but long-term safety has not been established. Endoscopic papillary balloon dilation is an alternative technique to sphincterotomy for removal of common bile duct stones (47). This sphincter-preserving technique is appealing for younger patients for whom the long-term risks of sphincterotomy remain unknown. However, published experience with this technique in children is scant (44), and the risk of pancreatitis may be higher than after sphincterotomy (48). Strictures associated with congenital and acquired lesions in the biliary system have been successfully treated in children by transendoscopic balloon dilation and placement of endoprostheses or stents (14,49–52). Biliary endoprostheses may also be used to treat acute or chronic bile leaks or fistula after abdominal trauma or surgical injury. Pancreatic Indications Pancreatic indications for diagnostic and therapeutic ERCP are listed in Table 3. ERCP may be used to identify intrinsic (e.g., anomalous pancreatobiliary junction) and extrinsic (e.g., gallstones) causes of persistent or recurrent acute pancreatitis, to evaluate complications of pancreatitis, and to treat mechanical obstruction or fluid collections. The utility of ERCP for investigating pancreatitis in children has been reported (53–57), and recent studies (7,12)(15,58) show abnormal findings in most children investigated for presumed idiopathic recurrent pancreatitis.TABLE 3: Pancreatic indications for ERCPDevelopmental anomalies identifiable by pancreatography have been associated with recurrent pancreatitis. The most common of these is pancreas divisum or so-called dominant dorsal duct (59), which is found in 5% to 10% of the general population. Endoscopic sphincterotomy of the minor papilla has led to improvement in approximately 75% of adult patients with recurrent acute episodes of pancreatitis associated with pancreas divisum (60), comparable with the outcome after surgical sphincteroplasty. Improved outcome has been reported after endoscopic therapy in children, but the data are limited to a small number of case reports (7,14,15,58,61,62). Anomalous union of the pancreatobiliary ductal system (63), annular pancreas (15), dorsal pancreatic agenesis (15), and duodenal duplications (64,65) have also been associated with recurrent episodes of pancreatitis in children and may be demonstrated by endoscopic pancreatography. Gallstone or biliary pancreatitis occurs in children and may necessitate endoscopic intervention. Based on studies in adult patients (66–68), early ERCP (within 24–48 hours of occurrence of symptoms) is indicated in acute biliary pancreatitis when a stone is identified in the common bile duct or when biliary obstruction or cholangitis is evident. Early endoscopic intervention for biliary drainage reduces the morbidity of suppurative cholangitis without significantly affecting the course of pancreatitis in adult patients (68,69). Similar outcome studies in children have not been reported. ERCP provides a valuable role in assessing integrity of the pancreas after abdominal trauma (70). Transection of the duct may be detected early and treated surgically or even endoscopically by placement of a transpapillary stent (61). When pseudocysts fail to resolve spontaneously after an episode of pancreatitis, ERCP may provide useful diagnostic information for planning surgical drainage. Endoscopic drainage may also be undertaken at the time of initial ERCP. Endoscopic options include: nasocystic drainage catheter, transpapillary endoprosthesis (stent), transmural cyst-duodenostomy or cyst-gastrostomy, and combination therapy with transmural and transpapillary drainage. Definitive therapy has been achieved in the majority (76%) of adult patients with an acceptable complication rate of 15%(71–76). Successful endoscopic drainage has been reported in children (7,15,58,61), but experience is insufficient to comment on either the efficacy or safety of this procedure in children. ERCP is indicated in the evaluation and treatment of chronic pancreatitis; direct pancreatography provides detailed information about anatomic changes in the main pancreatic duct or its side branches that may be used to confirm the diagnosis. The major indication for therapeutic ERCP in patients with chronic pancreatitis is the presence of stones within a dilated pancreatic duct. Recent adult studies (77,78) suggest that 80% to 100% of patients with either partial or complete clearance of intraductal stones enjoy immediate and complete relief of pain. As with other therapeutic interventions, pediatric experience is limited to a few case reports (15,61). In the absence of intraductal stones, endoscopic sphincterotomy may be considered as an alternative to surgical sphincteroplasty in the presence of obstructive findings, including intermittent or persistent ductal dilatation or abnormally prolonged dilatation with provocative testing (e.g., secretin stimulation) (79). Outcomes of this procedure in pediatric patients have not been established. CONCLUSION ERCP is a valuable diagnostic and therapeutic modality in carefully selected children with known or suspected hepatobiliary or pancreatic disease. Indications for ERCP in children are similar to those in adult patients. An experienced endoscopist, with advanced training in biliary and pancreatic techniques, should perform the procedure in a medical environment optimal for the care of children. When a pediatric gastroenterologist with such training and experience is not available, the procedure should be performed applying the combined knowledge and expertise of a pediatric gastroenterologist and an experienced endoscopist, usually an adult gastroenterologist. Based on the limited data available, the safety and efficacy of therapeutic interventions are comparable in children and adults. Acknowledgments: Members of the Subcommittee on Endoscopy and Procedures of the North American Society for Pediatric Gastroenterology and Nutrition include: Marsha Kay, M.D., Chair; Dennis Christie, M.D.; William Cochran, M.D.; Victor L. Fox, M.D.; Mark Gilger, M.D.; Aubrey Katz, M.D.; Seji Kitagawa, M.D.; Eric Maller, M.D.; Allan Rosenberg, M.D.; and Cory Strobel, M.D.