In this prospective cohort study with 1124 ERCPs (n=857 patients), 92% were performed in 15 centers by a pediatric gastroenterologist. Key findings:
Common indications included choledocholithiasis (41%), improved drainage for chronic pancreatitis (14%), stricture (17%), plan for sphincterotomy (5%) and bile leak (3%)
Procedures were technically successful in 90.5%
Only 26 cases (2.3%) were completed in infants (<1 yr) and 35 cases (3.1%) in 1-3 yrs. In these age groups, technical success was much lower: 80% (all children 3 yrs of age or less). Similarly, technical success was 76% in those <10 Kg.
ASGE difficulty grades 1 & 2 had success rates of 94% compared to 86% for grades 3 & 4. (ASGE classification adds one level in those 3 years of age or younger). ASGE grade 1, 2, 3 & 4 accounted for 13%, 47%, 34%, and 5% respectively.
Adverse effects were reported in 9.5%, though most were mild. Three perforations and 5 bouts of cholangitis were reported; however, there were no deaths or serious morbidity reported.
Post-ERCP pancreatitis (PEP) prophylaxis was associated with a decreased odds of PEP (OR 2.1, P=<0.01); among specific PEP prophylaxis, only rectal indomethacin neared statistical significance (P=0.07)
My take: Well-trained pediatric gastroenterologists are capable of doing high-quality ERCPs. In very young children, technical success is more difficult and probably requires the highest skillset.
Methods: This retrospective chart review (n=298) examined the outcomes of pediatric patients who receivied ketorolac during ERCP compared to those who did not; ketorolac, was dosed using weight-based dosing (0.5 mg/kg/dose max 30 mg).
Most common indications for ERCP were choledocholithiasis, biliary stricture, and chronic/recurrent pancreatitis
Therapeutic ERCP: 91% of ketorolac group and 89% of non-ketorolac group; sphincterotomy was performed in 55% of both groups
Post-ERCP bleeding rates were not significantly different between ketorolac and non-ketorolac groups (0.6% vs 0%, P = 1)
PEP (post-ERCP pancreatitis) rates were not significantly different between the ketorolac and no ketorolac group 15/166 vs 17/132 (9% vs 13%, P = 0.29)
Patients who had cannulation and/or injection of the PD had significantly higher rates of PEP (23/140 (16%) vs 9/158 (6%), P < 0.003)
For high-risk pediatric patients with injection of contrast into and/or cannulation of the pancreatic duct, the rates of PEP were significantly lower for patients who received ketorolac (11% vs 25%, P = 0.035). In Table 2, the authors indicate that PEP in this high risk group occurred in 11/88 (12.5%) [mild discrepancy from abstract of 11%] of ketorolac group and 13/52 (25%)
It is possible that ketorolac (or other NSAIDs) in all patients may be beneficial but difficult to demonstrate without a larger cohort. In adults, “indomethacin reduces risk and severity of PEP in both high and average risk adults…and there is some evidence that NSAIDS given before ERCP may be more effective than those given later.” Thus, the authors state that use of ketorolac could be administered to all ERCP patients beforehand (w/o contraindication) or limited to the higher risk patients.
My take: In pediatric patients needing an ERCP, those with high-risk features (eg. injection of contrast into and/or cannulation of the pancreatic duct), use of ketorolac is likely to reduce the frequency of post-ERCP pancreatitis.
A recent retrospective study (DS Fishman et al. JPGN 2020; 71: 203-207) identified 25 patients who underwent combined Laparascopic cholecystectomy/ERCP in Same Session (=LESS) to more conventional ERCP followed by laparoscopic cholecystectomy (n=42). The center utilized prospectively-collected data from 13 centers and 67 consecutive ERCPs.
Median hospital stay was shorter for LESS patients, 3 days vs. 4 days (P=.32)
Total procedure time was similar, though a decrease in total anesthesia time was reported for LESS patients: mean 177 minutes compared to 205 minutes (P=.04)
No significant adverse events were reported in either group, though both groups had two patients who required repeat care due to suspected retained stones
The authors note that concerns about gaseous distention following ERCP “is likely unfounded as all cholecystectomies were completed.”
No local or systemic infections were reported. The authors recommend antibiotic prophylaxis with the LESS approach
My take: Given the recommendation that cholecystectomy should take place during the same hospitalization for patients with choledocholithiasis, this combined approach makes a lot sense and is supported by this study.
Like the last two days, this post addresses “high quality” colonoscopy…
A recent report (M Thomson et al. JPGN 2019: 69: 6-12) describes quicker pediatric colonoscopy times than previously reported and with 100% rate of ileal intubation.
In this report, there were 181 colonoscopies.
Time of colonoscopy:
The authors emphasize the fact that their mean time to the terminal ileum was 9.8 minutes. Their good technical skill is probably related in part to experience: all 6 endoscopists had more than 10 years of experience (mean 19 years) and more than a thousand prior colonoscopies each.
Ileal Intubation Rate:
The 100% ileal intubation rate similarly indicates good technical skill. It may indicate that their patient population was healthier as ileal structuring (which can prevent ileal intubation) can be noted in patients with Crohn’s disease.
Low Diagnostic Yield:
In my view, the study reports a low diagnostic yield. They report 33% had abnormal histology (when excluding patients with IBD followup examinations)
38% of their patients had colonoscopy due to abdominal pain. They reported a yield in this group of only 11.6% though this includes 4 patients with “TI lymphoid hyperplasia.” Is this an abnormal finding?
My take: This study shows that with good technical skill colonoscopy can be done quickly with ileal intubation times averaging 10 minutes and with ileal intubation rates close to 100%. In my view, another quality metric is diagnostic yield and their yield is lower than has been reported in most pediatric studies.
K Siau et al. JPGN 2019; 69: 18-23. This study describes “Direct Observation of Procedural Skills” (DOPS). Among 29 trainees, 81% of DOPS were rate competent after 125-140 procedures.
MT Barakat et al. JPGN 2019; 69: 24-31. This study noted that the vast majority of pediatric GI centers (>90%) were performing less than 25 ERCPs annually and that >70% “believe their institution’s current arrangement for performing pediatric therapeutic endoscopy is inadequate.”
Many times throughout the year we will receive a request to accept a 15-17 year old weighing more than 200 pounds with gallstones who needs to be transferred so that he/she can be cared for in a pediatric facility. The really crazy part is that some of these ‘kids’ need to transferred back to an adult facility to have an ERCP to remove the gallstones if they are lodged in the common bile duct (CBD). Very few pediatric gastroenterologists are adequately trained in ERCP.
A recent retrospective study (PC Bonasso et al JPGN 2019; 68: 64-7) shows some of the consequences of this problem –longer hospitalizations and delays in treatment. The authors compared 79 (48%) pediatric patients who required transfer compared to 85 (52%) who were managed at the tertiary care pediatric hospital. The median age was 15 years with 42% obese and 23% overweight.
Transfer group patients had longer length of stay, median 7 days vs 5 days for non-transfer group (P< 0.0001) and more days between ERCP and surgery.
Transfer patients were more likely to have an MRCP (34% vs 8% for non-transfer).
Transfer patients were more likely to have a stent placement, 9% vs 5% (which would require a subsequent anesthetic to remove).
Transfer patients were more likely to have a non-therapeutic ERCP; stone/sludge removal was 70% in transfer group vs 86% in non-transfer group. This could be related to the delay (eg. more time for stone to pass) or due to the evaluation by team not responsible for ERCP.
The authors note that there are fewer than 20 pediatric gastroenterologists trained in ERCP; this is not likely to change much in the near term due to the large number of ERCPs needed to become proficient and few options for pediatric training. Their study notes that 46% had adult gastroenterologist management for non-transfer group.
My take: This is clearly an area in need of collaboration. More pediatric hospitals need to have adult gastroenterologists available and adult hospitals need to consider keeping some of these young adults to improve both the care and costs for these individuals.
An interesting study and accompanying editorial (RN Keswani et al. Clin Gastroenterol Hepatol 2017; 15: 1866-75, & P Cotton Clin Gastroenterol Hepatol 2017; 15:1855-57) point out that ERCP is more successful in high volume centers and with high volume (HV) endoscopists.
The study was a systematic review and meta-analysis. The threshold for low volume for endoscopist was < 27 case/year and for centers of <156 cases/year. However, this data is not widely available.
In this study with 59,437 ERCPs, HV endoscopists had OR of 1.6 for success compared to LV endoscopist.
Similarly HV centers had OR of 2.0 for successful ERCP.
Post-ERCP adverse risks were lower for HV endoscopists with OR of 0.7
In addition, the level of complexity for the cases matters a lot. Dr. Cotton breaks down the complexity of procedures:
Standard complexity includes cannulation of bile duct, biliary stent removal/exchange, biliary stone removal <10 mm, treating bile leaks, treating benign and malignant strictures, and placing prophylactic stents..
Advanced procedure complexity includes any of the above procedures after Billroth II, minor papilla cannulaiton, removing biliary stents that have migrated internally, fine needle aspiration, treating pancreatic strictures, removing small (< 5 mm), mobile pancreatic stones, treating biliary strictures at hilum or more proximal
Highly technical complexity (“advanced tertiary”) includes removal of internally migrated pancreatic stents, intraductal image-guided therapy (eg. PDT EHL), pancreatic stones impacted or >5 mm, intrahepatic stones, pseudocyst drainage, ampullectomy, ERCP after Whipple or roux-en-Y bariatric surgery
My take: The ultimate goal is high success rates and lower complication rates. Highly proficient endoscopists and high volume centers achieve these goals more consistently, particularly for more complicated ERCP procedures.
When parents make up diseases in their children, the consequences can be dire. What happens when doctors find out that a disease that they have been treating probably doesn’t exist?
Cotton PD et al. “Effect of endoscopic sphincterotomy for suspected sphincter of Oddi dysfunction on pain-related disability following cholecystectomy: the EPISOD randomized clinical trial” JAMA 2014; 311: 2101-2109.
A detailed analysis of this study (Gastroenterol 2015; 148: 440-44) and the author’s reply provides some insight into that questions and helps place this study and its results into context.
Key points from the Gastroenterology Selected Summary:
“Sphincter of Oddi dysfunction (SOD) is the term used to describe an episodic abdominal pain syndrome, typically occurring in young to middle-aged women in the setting of prior cholecystectomy…SOD remains a diagnosis of exclusion.”
“The investigators conducted a double-blind, sham-controlled, randomized trial at 7 US tertiary centers, enrolling 214 adult post-cholecystectomy patients with debilitating abdominal pain due to suspected biliary SOD (predominantly type III).” Sham patients underwent ERCP, manometry, and pancreatic duct stenting.
Patients were randomized 2:1 to sphincterotomy or sham; those randomized to sphincterotomy and subsequently shown to have a hypertensive pancreatic sphincter were then re-randomized (1:1) to have biliary or combined biliary/pancreatic sphincterotomies.
“Most patients in both study groups experienced considerable reduction in their pain disability scores…the proportion meeting the trial’s 1-year primary endpoint was higher among those treated with sham compared with sphincter ablation (37% vs. 23%, P=.01)”
“The manometry findings did not predict treatment success.” There were no other useful predictors of success identified (eg. elevated liver enzymes, prior stone at cholecystectomy)
Adverse effects from procedures included 26 cases of acute pancreatitis (2 severe) and 2 perforations.
“The results of this trial are fascinating…the authors, many of whom had dedicated entire careers to the management of these patients using the very procedure they have now conclusively shown to be futile, may be understandably disheartened by the results.”
Numerous limitations of the study are noted. In particular, “the 1-year time frame of the trial is likely to have been too short to capture the deleterious impact of prophylactic pancreatic duct stenting, which…has been associated with interval induction of pancreatic ductal abnormalities mimicking chronic pancreatitis.”
“It is our view that the authors’ data…provide an unambiguous mandate for imposing an immediate moratorium on subjecting this group of patients to ERCP.“
Dr. Cotton’s reply:
“It may be premature to discard the whole concept of sphincter dysfunction as a cause of pain.”
He indicates that “gallbladder dyskinesia” is another related question and was the reason for surgery in half of EPISOD subjects.
The results of the study “clearly show the need for equally stringent studies to answer the many remaining questions.”
For those who read this much of this post: I wanted to let you know that yesterday’s online post on early peanut introduction was updated with recommendations from the associated editorial.
While there are pediatric patients who undergo endoscopic retrograde cholangiopancreatography (ERCP), this is a relatively infrequent occurrence. Nevertheless, a recent study has a couple useful clinical pearls that may have broader application.
Aggressive hydration may prevent post-ERCP pancreatitis. In the study, the treatment group received an average of 3290 mL over the 9-hour period compared with 945 mL in the standard infusion group.
Implication: The speculation from the study and the editorials is that improved pancreatic perfusion will result in better oxygenation and reduce the likelihood of pancreatitis. In the 2nd reference, the author states that his practice is to administer “at least 3 L of crystalloid in recovery to young, healthy patients who have undergone high-risk ERCP and an additional 3 to 5 L within the first 12 hospital hours to those admitted with postprocedure pain”
The best fluid (for post-ERCP and acute pancreatitis) may be lactated Ringer’s (LR).
Implication: The lactate in LR may help reduce pancreatitis by avoiding acidosis which could promote zymogen activation and pancreatic inflammation. A previous small trial (n=40) of acute pancreatitits from any cause showed lesser degrees of systemic inflammatory response with LR in compared with normal saline (Clin Gastroenterol Hepatol 2011; 9: 710-17e1).
This study adds aggressive IVFs as another intervention to prevent ERCP. Rectal indomethacin and prophylactic stent placement (in high-risk patients) are other accepted treatments.
This pilot study randomly assigned 39 patients to aggressive hydration and 23 to standard hydration; all patients were inpatients who were not at risk for fluid overload. The aggressive group received 3 mL/kg/h during the procedure, a 20 mL/kg bolus after the procedure, and then continued on 3 mL/kg/hr for 8 hours. In contrast, the standard group received LR at 1.5 mL/kg/h during and for 8 hours afterwards.
Demographics: The average age was 43 years in the aggressive hydration group and 45 years in the standard group. 78% were hispanic. The ERCP procedures were mostly “average risk.” 74% had ERCP for choledocholithiasis. Only 2 subjects needed precut sphinterotomy (3%).
No patients in the aggressive hydration group developed acute pancreatitis compared with 4 (17%) in the standard hydration group
Elevated amylase (23% vs. 39%) and epigastric pain (8% vs 22%) were also less frequent in the aggressive hydration group.
Numerous Limitations: This was a small pilot study with an atypical population; thus, the findings are difficult to generalize. A false-positive (type 1 error) can easily occur due to the small numbers, especially as the standard hydration group had a rate of acute pancreatitis that was about double from previous studies. In addition, this study was not blinded and could have been susceptible to bias. Furthermore, the authors defined acute pancreatitis differently than in previous studies. In this study, the authors required enzyme increases 2 or 8 hours after ERCP with new abdominal pain; in previous studies, the definition of acute pancreatitis relied on enzyme increases for at least 24 hours after the ERCP.
Take-home message for those not doing ERCPs: Think about using lactated ringer’s and aggressive hydration in otherwise-well patients who present with acute pancreatitis.
A recent case vignette highlights several key points regarding use and timing of ERCP (endoscopic retrograde cholangiopancreatography) for gallstone pancreatitis (NEJM 2014; 370: 150-7). Figure 1 provides a nice illustration of ERCP.
Indications: Suspected bile-duct stones as the cause of pancreatitis AND one of the following:
“Take two and call me in the morning” may now apply to the use of indomethacin in preventing post-ERCP pancreatitis. A multicenter, randomized, placebo-controlled, double-blind clinical trial has shown that rectal indomethacin (two 50 mg suppositories) can reduce the rate of post-ERCP pancreatitis (NEJM 2012; 366: 1414-22).
A total of 602 adult patients were enrolled. Patient selection favored those at increased risk for post-ERCP pancreatitis (eg. suspicion of sphincter of Oddi dysfunction) and excluded those at low risk for this complication (eg. routine biliary stent exchange, chronic calcific pancreatitis, or a pancreatic head mass). Other exclusion criteria included active pancreatitis, elevated creatinine (>1.4 mg/dL), active peptic ulcer disease, and those already receiving a NSAID.
The suppositories (or placebo) were administered immediately after ERCP while the patient remained in the procedure room.
Post-ERCP pancreatitis developed in 27 of 295 patients (9.2%) in indomethacin group and in 52 of 307 patients (16.9%) in placebo group. In addition, moderate-to-severe pancreatitis was reduced as well 4.4% compared with 8.8% respectively. In addition, there were no increased adverse events in the treatment arm; there was no increased risk of bleeding in particular.
While the mechanism of improvement is unclear, NSAIDs are potent inhibitors of phospholipase A2, cyclooxygenase, and neutrophil-endothelial interactions, all of which are known to play a role in the pathogenesis of acute pancreatitis.
-Am J Gastroenterol 2007; 102: 978-83. Use of indomethacin to reduce pancreatitis after ERCP
-Gut 2008; 57: 1262-7. Meta-analysis of rectal NSAIDs to prevent post-ERCP pancreatitis