Post–Endoscopic Retrograde Cholangiopancreatography Pancreatitis

Post-ERCP pancreatitis (PEP) remains the most common and most feared adverse outcome in the realm of therapeutic endoscopy. The reported incidence of PEP for diagnostic and therapeutic ERCP is 0.4% to 1.5% and 1.6% to 5.4%, respectively. High-risk cohorts (patients with sphincter of Oddi dysfunction [SOD] undergoing ERCP) have a much higher rate of PEP.

Patient-related factors have been extensively evaluated, and include suspected or known SOD (raising the rate to 10%–30% ) as well as prior PEP (18%–26%) A normal serum bilirubin also carries an increased risk of PEP. Previous pancreatitis, even when not due to an ERCP, also increases the risk of PEP.

Young patient age (usually considered <30 years) is associated with an increased risk of PEP. Female gender is a well-established risk factor for PEP, with a relative risk of 2.23. Current and former alcohol users/abusers, along with former smokers, have a 2- to 3-fold risk for PEP. The presence of pancreas divisum increases the risk of PEP to 8.2% when dorsal pancreatic duct cannulation is attempted (even if not successful). Multiple high-risk factors seem to have a cumulative increase in the risk of developing PEP (ie, a young woman with presumed SOD and normal serum bilirubin).

Conversely, there are a few patient-related or disease-related factors that seem to impart a lower rate of PEP. For example, elderly patients experience a lower rate of PEP when compared with younger patients, and these patients tolerate ERCP well overall. A study looking at combined ERCP and EUS in elderly patients showed that this approach was considered safe and effective, even in very elderly patients, and even when taking into account the increased length of the procedure and the need for deep sedation/general anesthesia. Patients with pancreatic solid tumors (most commonly pancreatic adenocarcinoma of the head) are at decreased risk for PEP when undergoing purely biliary procedures, owing to chronic obstruction of the pancreatic duct along with atrophy of the upstream pancreatic parenchyma.

With regard to procedure-related factors, several are known to increase the risk of PEP. The so-called difficult cannulation is a nonspecific term that is used to refer to a failure to obtain biliary access in a rapid or straightforward manner, and typically implies multiple and/or prolonged attempts at achieving deep biliary cannulation (whether or not cannulation was actually achieved). The term often implies that a combination of factors may have occurred during an attempt at cannulation, including the possibilities of associated mucosal trauma to the ampulla from repeated cannulation attempts, use of the needle-knife sphincterotomy technique, and/or multiple undesired pancreatic duct injections. Patients with a difficult cannulation are at an elevated risk of developing PEP.

Needle-knife sphincterotomy is an established independent risk factor for PEP. Some of the risk related to needle-knife sphincterotomy seems to be because a large number of cannulation attempts performed before the technique is used in most cases, arguing for the acceptability of early needle-knife attempts if standard approaches fail. Some of the risk is also due to the nature of the needle-knife sphincterotomy itself (ie, cautery to access the bile or pancreatic duct without a guide wire to ensure proper access and orientation when making the sphincterotomy). A retrospective study by Testoni and colleagues showed that the rate of PEP is lower when needle-knife sphincterotomy was initiated after fewer than 10 cannulation attempts, which many would consider a very early transition to this technique (ie, most physicians would likely continue to attempt to achieve cannulation by standard techniques after 10 failed attempts). A 2010 meta-analysis by Cennamo and colleagues demonstrated PEP in 2.5% of patients randomized to early needle-knife sphincterotomy, compared with 5.3% of patients undergoing persistent cannulation attempts before needle-knife sphincterotomy, also arguing that earlier application of this technique may be warranted over continued attempts at cannulation with standard devices, although this conclusion has not been embraced by the wider GI community.

Endoscopic papillary balloon dilation (EPBD) without prior biliary sphincterotomy has been associated with a high rate of PEP and adverse outcomes in Western patients, although it is much more commonly performed in Asia, most notably Japan. In a well-publicized, randomized controlled trial, EPBD without sphincterotomy was associated with a significantly increased risk of PEP, and the study included 2 deaths from pancreatitis occurring in patients who underwent EPBD without biliary sphincterotomy. PEP is widely known to be associated with repeated pancreatic duct injections.

Indomethacin as a prophylactic agent is widely being used in patients undergoing ERCP, based on a recent New England Journal of Medicine study showing a reduced incidence of PEP in those receiving 1 dose of rectal indomethacin at the conclusion of their ERCP. In this study, most patients (82%) were considered at high risk for PEP because they had suspected SOD. Therefore, the use of indomethacin in all patients undergoing ERCP as opposed to only selected high-risk patients is currently under debate. However, given the ease and safety of a single dose regimen and its low cost, the routine use of rectal indomethacin as a PEP-preventive measure is gaining popularity.

Perforation

Risk factors that increase the potential for perforation to occur during endoscopy have been identified, and physicians should be aware of the more common ones. Endoscopic dilation of strictures (especially in patients with inflammatory bowel disease, most notably Crohn disease) or large-caliber dilation in achalasia are considered to be higher-risk procedures with regard to perforation. Furthermore, the performance of endoscopic mucosal resection or endoscopic submucosal dissection should also be considered a relatively high-risk procedure with regard to transluminal perforation in comparison with standard endoscopy. Chronic steroid use can also increase the risk of perforation during any endoscopic procedure.

In patients undergoing ERCP, perforation is most likely to occur during sphincterotomy, but can also occur during intubation of the patient at the esophagus, stomach, or duodenum. Perforations may be more common in patients with postsurgical anatomy. Most sphincterotomy-related perforations are retroperitoneal and, if recognized, can often be managed conservatively with nil by mouth status, antibiotics, and close monitoring. It should be mentioned that small retroperitoneal perforations might be difficult to identify clinically during ERCP.

Small bowel enteroscopy, once an uncommon procedure performed mostly at tertiary referral centers, is now in widespread use in both academic and community practices, as a result of the dissemination of several new endoscopic techniques (eg, single-balloon enteroscopy, double-balloon enteroscopy, spiral enteroscopy) that are now available to practitioners to allow them access to the entire small bowel, in either an anterograde or retrograde fashion. Perforations during these types of enteroscopy procedures are more likely to occur during interventions such as the treatment of bleeding, strictures, and when attempting to remove large polyps, although perforations can occur during advancement or removal of the endoscope, especially in patients with postsurgical anatomy.

In patients with inflammatory bowel disease, patient-related risk factors associated with higher rates of perforation include severe/active inflammatory bowel disease, active or long-term corticosteroid use, female sex, older age, and the performance of endoscopic dilations (se earlier discussion).

Elderly patients (variously defined as >60, 65, or 70 years) have been noted to have an increased risk of perforation, as have patients with multiple comorbidities. Elderly patients can also have an increased risk of cardiopulmonary complications, and these must be addressed before the procedure, especially when planning for the appropriate type of anesthesia to administer and the need for additional anesthesia support. Other factors that increase the risk of perforation include extensive diverticular disease, multiple abdominal or pelvic surgeries (which can cause fixation of the colon resulting from adhesions), and bowel obstruction.

Historically, transmural perforation of the bowel at almost any location mandated surgery to repair, but this dogma is rapidly changing. The increasing use of endoscopic clipping has allowed many of these perforations to be directly closed endoscopically. These clips work best if the perforation is focal, well visualized, and if there has been a paucity of or no soiling of the abdomen with enteric contents. Through-the-scope or over-the-scope clips are widely available and are highly effective for treating small or localized perforations, but may fail to close perforations with a large transmural defect. Endoscopic suturing devices are also available. Large perforations, perforations that cannot be seen well, or patients who have soiling of the abdomen with copious amounts of enteric contents will likely need surgery to repair these defects.

Bleeding

Bleeding, especially significant bleeding requiring transfusion or other interventions, is a rare event in modern endoscopy. Nonetheless, at some point all endoscopists are likely to cause a clinically significant bleed.