Post–endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis is a common and potentially devastating complication of ERCP. Advances in risk stratification, patient selection, procedure technique, and prophylactic interventions have substantially improved the endoscopists’ ability to prevent this complication. This article presents the evidence-based approaches to preventing post-ERCP pancreatitis and suggests timely research questions in this important area.
Key points
- •
Risk stratification and thoughtful patient selection are critical in preventing post-ERCP pancreatitis; in this era of highly accurate diagnostic alternatives, ERCP should be a near-exclusively therapeutic procedure.
- •
In the case of difficult cannulation, alternate techniques, such as double-wire cannulation and precut sphincterotomy, should be implemented early.
- •
Contrast-facilitated cannulation, aggressive/repeated pancreatic injection, dilation of an intact biliary sphincter, and sphincter of Oddi manometry without aspiration should be avoided.
- •
Prophylactic pancreatic stents should be placed in all high-risk cases.
- •
Rectal nonsteroidal antiinflammatory drugs should be administered in all high-risk cases and based on a very favorable risk-benefit ratio, should be considered in all patients undergoing ERCP.
Overview
Post–endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis (PEP) is defined as new or increased abdominal pain that is clinically consistent with a syndrome of acute pancreatitis, pancreatic enzyme elevation at least 3 times the upper limit of normal 24 hours after the procedure, and resultant hospitalization (or prolongation of existing hospitalization) by more than 1 night. Pancreatitis is still the most common complication of ERCP, occurring in 2% to 10% of cases and accounting for substantial morbidity, occasional mortality, and health care expenditures in excess of $150 million annually in the United States. Despite substantial advances over the last several decades in patient selection, equipment, procedural technique, and prophylactic interventions, PEP remains a serious health problem, and its prevention remains a major clinical and research priority. Herein is an evidence-based review of approaches to prevent pancreatitis after ERCP, as well as suggestions for necessary research objectives in this important area.
Overview
Post–endoscopic retrograde cholangiopancreatography (ERCP) pancreatitis (PEP) is defined as new or increased abdominal pain that is clinically consistent with a syndrome of acute pancreatitis, pancreatic enzyme elevation at least 3 times the upper limit of normal 24 hours after the procedure, and resultant hospitalization (or prolongation of existing hospitalization) by more than 1 night. Pancreatitis is still the most common complication of ERCP, occurring in 2% to 10% of cases and accounting for substantial morbidity, occasional mortality, and health care expenditures in excess of $150 million annually in the United States. Despite substantial advances over the last several decades in patient selection, equipment, procedural technique, and prophylactic interventions, PEP remains a serious health problem, and its prevention remains a major clinical and research priority. Herein is an evidence-based review of approaches to prevent pancreatitis after ERCP, as well as suggestions for necessary research objectives in this important area.
Recognizing patients at increased risk for PEP
PEP prevention begins with recognition of patients at increased risk, because a high index of suspicion for and early identification of post-ERCP pancreatitis are critically important in ensuring favorable clinical outcomes. The ability to risk stratify patients based on well-established clinical characteristics can concretely influence the decision-making process that surrounds PEP prevention and the management of its potentially devastating sequelae. Armed with the risk assessment information outlined later in the discussion, clinicians can tailor costly and potentially dangerous risk-reducing strategies. For example, prophylactic pancreatic stent placement (PSP) and consideration of post-procedure hospital observation are appropriate for a patient predicted to be at high risk for PEP but are not justified in low-risk cases.
A substantial amount of research over the last 2 decades has contributed to the understanding of the independent risk factors for post-ERCP pancreatitis. These risk factors, listed in Table 1 , are divided into patient-related and procedure-related characteristics. The definite and probable patient-related risk factors that predispose to PEP are a clinical suspicion of sphincter of Oddi dysfunction (SOD) (regardless of whether or not sphincter of Oddi manometry is performed) ; a history of prior PEP ; and a history of recurrent pancreatitis, normal bilirubin, younger age, and female gender. The definite and probable procedure-related risk factors for PEP are difficult cannulation, precut (access) sphincterotomy, pancreatic sphincterotomy, ampullectomy, repeated or aggressive pancreatography, balloon dilation of an intact biliary sphincter, and possibly passage of a guidewire deep into the pancreatic duct (PD). An important risk-stratification principle is that predictors of PEP appear synergistic in nature. For example, a widely referenced multicenter study by Freeman and colleagues, predating prophylactic PSP, showed that a young woman with a clinical suspicion of SOD, normal bilirubin, and a difficult cannulation has a risk of PEP in excess of 40%. Of note, biliary sphincterotomy and Billroth 2 anatomy do not appear to predispose to PEP.
| Patient-related Factors | Procedure-related Factors |
|---|---|
| Suspected sphincter of Oddi dysfunction (SOD) | Difficult cannulation |
| Prior post-ERCP pancreatitis | Precut (access) sphincterotomy |
| Normal bilirubin | Pancreatic sphincterotomy |
| Younger age | Ampullectomy |
| Female gender | Repeated or aggressive pancreatography |
| History of recurrent pancreatitis | Balloon dilation of an intact biliary sphincter |
Patient-related characteristics are not modifiable but can be used (at least in part) to predict the risk of PEP before ERCP, allowing appropriate case selection and a meaningful discussion with the patient regarding the risk-benefit ratio of the procedure. For example, a young woman with suspected biliary SOD but moderate symptoms that are partially responsive to pain-modulating therapy may elect to forgo ERCP after understanding her elevated risk of severe PEP.
Procedural risk factors may occasionally be modified during the case (see later discussion) but in combination with patient-related factors, allow a global assessment of a patient’s overall risk profile, guiding the implementation of appropriate prophylactic interventions, such as aggressive intravenous fluid administration, rectal indomethacin, PSP, and a lower threshold for hospital observation after the procedure.
Operator (endoscopist)-dependent characteristics have also been implicated in the risk of PEP. Endoscopist procedure volume is suggested to be a risk factor for PEP, although multicenter studies have not confirmed this trend, presumably because low-volume endoscopists tend to perform lower-risk cases. Nevertheless, potentially dangerous cases (based on either patient-related factors or anticipated high-risk interventions) are best referred to expert medical centers where a high-volume endoscopist with expertise in prophylactic PSP can perform the case and where more experience with rescue from serious complications may improve clinical outcomes. Similarly, trainee involvement in ERCP is a possible independent risk factor for PEP, although results of existing multivariable analyses are conflicting. It stands to reason that inexperienced trainees may augment procedure-related risk factors, such as prolonging a difficult cannulation or delivering excess electrosurgical current during an inefficient pancreatic sphincterotomy, etc. Therefore, an improved understanding of the process of ERCP training is necessary to minimize the contribution of trainee involvement to the development of PEP. Future research focused on defining ERCP training metrics and developing an evidence-based list of appropriate fellow cases based on stage of training and skill level is needed. Further, defining the optimal parameters that guide trainee-attending scope exchange during any particular case or intervention is necessary to maximize learning potential while minimizing patient risk.
Several additional points regarding clinical risk stratification are worth considering. First, patients with a clinical suspicion of SOD, particularly women, are not only at increased risk for PEP in general, but are also more likely to develop severe pancreatitis and death. When considering the risk-benefit ratio of ERCP in this patient population, not only should the patient’s overall risk of PEP be assessed, but their probability of experiencing a more dramatic clinical course should also be considered and discussed. An additional consideration is that several clinical characteristics are thought to significantly reduce the risk of PEP. First, biliary interventions in patients with a preexisting biliary sphincterotomy probably confer a very low risk of PEP. Prior sphincterotomy will have generally separated the biliary and pancreatic orifices ( Fig. 1 ), allowing avoidance of the pancreas and making pancreatic sphincter or duct trauma unlikely. Further, patients with chronic pancreatitis, in particular those with calcific pancreatitis, are at low risk for PEP because of gland atrophy, fibrosis, and consequent decrease in exocrine enzymatic activity. Similarly, the progressive decline in pancreatic exocrine function associated with aging may protect older patients from pancreatic injury. Lastly, perhaps because of post-obstructive parenchymal atrophy, patients with pancreatic head malignancy appear to be relatively protected as well.
Although understanding these aforementioned risk factors and incorporating them into clinical decision-making are important aspects of preventing PEP, additional research focused on developing more robust risk-stratification tools based on existing literature and future multicenter studies is important. Such risk stratification instruments are unlikely to be developed using conventional statistical models (ie, multivariable regression analysis) but may require the use of novel, more advanced prediction methods involving artificial intelligence, such as machine learning, a technique that has already been successfully utilized in both business and medicine. In addition, a more specific understanding of how these tools’ output should concretely direct clinical management is necessary.
Patient selection
Thoughtful patient selection before ERCP remains a fundamental strategy for preventing pancreatitis. Endoscopic ultrasound (EUS) and magnetic resonance cholangiopancreatography (MRCP) allow highly accurate pancreaticobiliary imaging while avoiding the significant risks of ERCP. Two large meta-analyses have demonstrated that EUS is highly sensitive and specific in the detection of bile duct stones (sensitivity 89%–94%; specificity 94%–95%). Similarly, MRCP has a sensitivity of 85% to 92% and a specificity of 93% to 97% for the same indication, although magnetic resonance imaging (MRI) appears less sensitive than EUS for stones smaller than 6 mm. Additionally, EUS, MRI, and other noninvasive modalities such as radionucleotide-labeled scan and percutaneous drain fluid analysis are very accurate in diagnosing a multitude of other pancreaticobiliary processes (eg, chronic pancreatitis, malignancy, and leaks), often obviating the need for ERCP.
Indeed, the utilization of ERCP as a diagnostic procedure has steadily declined in favor of less invasive but equally accurate alternative tests, and ERCP has appropriately become a near-exclusively therapeutic procedure reserved for patients with a high pretest probability of intervention. This trend is consistent with recent clinical practice guidelines on the role of endoscopy in the evaluation of choledocholithiasis and the National Institutes of Health consensus statement on ERCP for diagnosis and therapy, both favoring less-invasive tests over ERCP in the diagnosis of biliary disease.
An exception to the widespread practice of reserving ERCP for patients with a high likelihood of therapeutic intervention is the evaluation of patients with suspected SOD, for which an accurate, less-invasive alternative to ERCP-guided sphincter of Oddi manometry (SOM) remains elusive. Even when considering patients for SOM, however, thoughtful clinical judgment is necessary to select those who are most likely to benefit from the procedure. An ongoing multicenter clinical trial (the EPISOD study) evaluating the role of ERCP in patients with suspected SOD but no laboratory or radiographic abnormalities will help clarify the risk-benefit ratio of ERCP for this population and hopefully allow improved patient selection in this challenging context. Another possible exception to the therapeutic ERCP trend may be the evaluation of biliary complications in liver transplant recipients, for whom a recent retrospective study suggested that diagnostic ERCP is a reasonable and efficient clinical approach in this patient population based on a high likelihood of therapeutic intervention and a very low rate of complications, in particular PEP.
Procedure technique
Efficient and atraumatic technical practices during ERCP are central to minimizing the risk of pancreatitis. Many of the procedure-related risk factors listed earlier, while predisposing to PEP, are mandatory elements of a successful case. Even though these high-risk interventions are unavoidable for execution of the clinical objective, certain strategies can be utilized to minimize procedure-related risk.
As mentioned, difficult cannulation and PD injection are both independent risk factors for PEP. As such, interventions that improve the efficiency of cannulation and limit injection of contrast into the pancreas are likely to decrease the risk of pancreatitis. Guidewire-assisted cannulation accomplishes both, representing a major paradigm shift in ERCP practice. In contrast to conventional contrast-assisted cannulation, which may lead to inadvertent injection of the PD or contribute to papillary edema, guidewire-assisted cannulation uses a small-diameter wire with a hydrophilic tip that is initially advanced into the duct, subsequently guiding passage of the catheter. Because the wire is thinner and more maneuverable than the cannula, it is easier to advance across a potentially narrow and off-angle orifice. Moreover, this process limits the likelihood of an inadvertent pancreatic or intramural papillary injection. A recent Cochrane Collaboration meta-analysis, which included 12 randomized controlled trials (RCTs) involving 3450 subjects, indeed confirms that guidewire-assisted cannulation reduces the risk of PEP by approximately 50% (relative risk [RR] 0.51, 95% confidence interval [CI] 0.32–0.82). A more recent prospective cohort study enrolling a mix of high-risk and low-risk subjects revealed no difference in PEP between the contrast and guidewire-assisted groups ; however, the results of this study have been questioned for a multitude of reasons, including the selection bias introduced by the nonrandomized design of the study. When wire cannulation is used for biliary access, it is important to advance the guidewire cautiously in the event in case it is actually in the PD where forceful advancement of the wire may induce pancreatitis by perforating a sidebranch.
When initial cannulation attempts are unsuccessful, several alternative techniques are available to facilitate biliary access. The double wire technique is a common second-line approach when initial cannulation attempts result in repeated unintentional passage of the wire into the pancreas. The wire can be left in the PD, thereby straightening the common channel and partially occluding the pancreatic orifice, allowing subsequent biliary cannulation alongside the existing pancreatic wire. The double wire technique has been shown to improve cannulation success compared with standard methods, although some data suggest a higher incidence of PEP with this technique or when a wire is passed inadvertently into the PD. Furthermore, a recent RCT of difficult-cannulation cases requiring double wire technique demonstrated that prophylactic PSP reduced the incidence of PEP by approximately 90% in this patient population. On this basis, some experts believe that a prophylactic pancreatic stent should be placed in all patients requiring double wire cannulation or when the wire inadvertently passes more than once into the pancreas. Others, including the authors, however, believe that placement of a wire in the pancreas does not independently predispose to PEP and that pancreatitis in this context is generally related to the preceding difficult cannulation. If the double wire technique is used early in a low-risk patient (within 2–3 cannulation attempts) and the wire advances seamlessly into the PD in a typical pancreatic trajectory, PSP may not be necessary.
Additional alternative cannulation techniques include wire cannulation alongside a pancreatic stent, precut sphincterotomy, septotomy, and needle-knife fistulotomy. Although these techniques are immensely helpful in gaining biliary access during challenging cases, some have been implicated as procedure-related risk factors for PEP. In many cases, however, the risk of PEP is actually driven by the preceding prolonged cannulation time that leads to increasing papillary trauma/edema. Therefore, implementing alternate cannulation techniques early in the case and in rapid succession is an important aspect of reducing PEP. This principle is best demonstrated by a recent meta-analysis of 6 randomized trials, which showed that early precut sphincterotomy significantly reduced the risk of PEP when compared with repeated standard cannulation attempts (2.5% vs 5.3%, odds ratio 0.47). However, the studies included in this meta-analysis were conducted in mostly low-risk patients, often with favorable anatomy for precut sphincterotomy. Further studies are needed to define the exact point at which the risk-benefit ratio favors precut sphincterotomy over repeated cannulation attempts, although the natural tendency to continue standard cannulation attempts beyond 5 to 10 minutes should be controlled, and alternative strategies should be attempted early in a difficult case.
Other technical strategies that reduce the risk of PEP include avoiding the frequency and vigor of PD injection, performing SOM using the aspiration technique and avoiding balloon dilation of an intact biliary sphincter, especially without prophylactic PSP. In coagulopathic patients with choledocholithiasis and native papillae, balloon dilation can be avoided by providing real-time decompression with a bile duct stent and repeating the ERCP with sphincterotomy and stone extraction when coagulation parameters have been restored. If this is not possible, and balloon dilation is mandatory, longer duration dilation (2–5 minutes) appears to result in lower rates of pancreatitis compared with 1-minute dilation. Of note is that balloon dilation after biliary sphincterotomy to facilitate large-stone extraction does not appear to increase the risk of PEP. All these factors are modifiable and should be considered during every ERCP.
Procedure equipment
Recent advances in ERCP equipment have increased technical success rates but have unfortunately not reduced the risk of post-ERCP pancreatitis. In particular, the use of a sphincterotome or a steerable catheter has been shown to improve cannulation success compared with a standard cannula but does not result in lower PEP rates. Similarly, comparative effectiveness studies evaluating sphincterotomes of various diameters have shown no difference in the risk of PEP. There are no comparative effectiveness data evaluating the effect of various guidewires on the risk of pancreatitis.
Along these same lines, the type of contrast medium used during pancreatography does not appear to affect the incidence of PEP, and it remains unclear (but unlikely) that the now commonly used automated electrosurgical current delivery systems offer any protection over the previously popular pure-cut current for thermal injury–induced pancreatitis.
Overall, it appears that equipment has little to no impact on post-ERCP pancreatitis. Therefore, practitioners should use the devices with which they are most comfortable for any particular indication to maximize technical success and efficiency, the latter of which is likely inversely related to the risk of PEP.
Prophylactic Pancreatic Stent Placement
One of many proposed mechanisms of PEP implicates impaired PD drainage caused by trauma-induced edema of the papilla. PSP is therefore thought to reduce the risk of PEP by relieving PD hypertension that develops as a result of transient procedure–induced stenosis of the pancreatic orifice. Eight RCTs (>650 subjects) and at least 10 nonrandomized trials have consistently demonstrated that PSP reduces the risk of PEP by approximately 60% to 80%. In the most recently published meta-analysis of RCTs, PSP resulted in an absolute PEP risk reduction of 13.3% (95% CI, 8.8%–17.8%) with a number needed to treat to prevent 1 episode of 8 (95% CI, 6–11). Equally importantly, prophylactic pancreatic stents appear to profoundly reduce the likelihood of severe and necrotizing pancreatitis.
It is important to keep in mind that the demonstrated benefits of PSP must be weighed against several potential disadvantages. First, attempting to place a PD stent with subsequent failure actually increases the risk of PEP above baseline by inducing injury to the pancreatic orifice but providing no subsequent ductal decompression. Second, significant nonpancreatitis complications induced by PSP, such as stent migration and duct perforation, occur in 4.4% of cases. Further, prolonged stent retention may induce ductal changes that resemble chronic pancreatitis, although the long-term clinical relevance of these changes remains unclear. PSP is associated with some patient inconvenience and increased costs by mandating follow-up abdominal radiography to ensure spontaneous passage of the stent and additional upper endoscopy to retrieve retained stents in 5% to 10% of cases.
Despite these considerations, PSP is widely regarded as an effective means of preventing PEP, is commonly used in academic medical centers in the United States, and is recommended by the European Society of Gastrointestinal Endoscopy. In light of the aforementioned concerns and the associated costs, however, PSP should be reserved for high-risk cases. Based on the known independent patient and procedure-related risk factors for PEP, experts have suggested that the following cases are appropriate for prophylactic PD stent placement: (1) clinical suspicion of SOD (whether or not manometry or therapeutic intervention performed), (2) prior PEP, (3) difficult cannulation, (4) precut (access) sphincterotomy, (5) pancreatic sphincterotomy (major or minor papilla), (6) endoscopic ampullectomy, (7) aggressive instrumentation or injection of the PD, and (8) balloon dilation of an intact biliary sphincter. Furthermore, a preliminary study has suggested that “salvage” PSP may be beneficial early in the course of PEP for patients who did not originally receive a stent or in the case of early stent dislodgement. Additional studies are necessary to fully evaluate PSP for this indication.
Several questions regarding optimization of the PSP process remain. First, there is limited consensus regarding the optimal stent length and caliber. An early study suggested improved outcomes with 3-F or 4-F stents ; however, a subsequent RCT showed no difference in PEP rates but a higher insertion success rate with the 5-F stents. Similarly, there is little consensus regarding optimal stent length. Most experts agree that the intrapancreatic tip of the stent should not rest at the pancreatic genu ( Fig. 2 ) ; however, whether short stents (ending in the pancreatic head) or longer stents (ending in the body or tail) are preferable is unknown, and comparative effectiveness studies in this area are needed.
Another important question regarding PSP is the acceptable amount of time that is spent on the insertion process in cases of difficult pancreatic access. Although the merits of PSP have been clearly presented earlier, if achieving pancreatic access proves difficult, there is presumably a point of diminishing returns when the risk of additional attempts outweighs the benefit of stent placement, especially if insertion eventually proves unsuccessful. Future clinical studies are unlikely to answer this question in a methodologically rigorous fashion; therefore, endoscopists should be aware of this important clinical balance and use their best judgment regarding the acceptable duration of time for stent insertion. One potential approach to circumvent this problem in cases of anticipated stent placement (for example ampullectomy or SOD cases) is to place and maintain a guidewire in the PD early in the case to guarantee PD access later on, avoiding the occasional phenomenon of failing to identify the pancreatic orifice due to the anatomic distortion that develops as a consequence of edema, sphincterotomy, or ampullectomy ( Fig. 3 ). Another approach is to place the prophylactic pancreatic stent before therapeutic intervention.
Related posts:
Advances in Pancreatic Endoscopy
Effective Endoscopic Management of Pancreatic Diseases
Role of Endoscopic Ultrasonography in the Diagnosis of Acute and Chronic Pancreatitis
Endoscopic Therapy for Chronic Pancreatitis
ERCP for Biliary Strictures Associated with Chronic Pancreatitis
Endoscopic Ultrasonography–Guided Drainage of the Pancreatic Duct
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
