Unexplained Acute Pancreatitis and Acute Recurrent Pancreatitis

Pathophysiology and Role of ERCP, EUS, and MRCP

Data suggest that the pathophysiologic process of AP might consist of three phases. The initial phase involves triggering events that are, for the most part, extrapancreatic in origin. Clinically, the most important of these appears to be either passage of a biliary tract stone or ingestion of ethanol. Other events, such as exposure to pancreatotoxins, pancreatic ischemia, and infection, may also be capable of triggering AP. The second phase involves a series of events that occur within the acinar cells of the pancreas. Finally, a third phase consisting of both acinar cell and nonacinar cell events occurs, which determines the ultimate severity of an attack of pancreatitis. There are two key points with these events that are important with regard to endoscopic intervention: (1) pancreatic ductal (PD) obstruction leads to ductal hypertension that is exacerbated by pancreatic secretion, and (2) ductal hypertension causes inhibition of enzyme secretion, resulting in colocalization of inactive pancreatic enzymes and lysosomal hydrolases with subsequent acinar cell injury and the clinical sequelae of AP. Several factors play an etiologic role in ARP; in fact, any cause of AP can lead to recurrent episodes if not corrected. Thus the role of ERCP, EUS, and MRCP in patients with IAP and IARP is to identify causes of triggering events that led to the PD obstruction, with the therapeutic goal of relieving the obstruction. It is assumed that relief of the obstruction will prevent further episodes of pancreatitis. The obstructive theory of IAP also assumes that ductal obstruction is intermittent or that a second risk factor predisposes patients with impaired ductal drainage.

Diagnostic Findings and Timing of ERCP, EUS, and MRCP

There are two major concerns that prompt physicians to do a more intensive evaluation of the patient with IAP. The first is that the patient may have an underlying disease that will predispose to further attacks of AP unless the cause is identified and adequately treated. AP is likely to recur in 33% to 67% of patients with biliary tract disease when not diagnosed and treated. Similarly, other anatomic or functional disorders of the pancreaticobiliary tree may predispose patients to recurrent episodes of pancreatitis. The second concern is that the pancreatitis may be related to a tumor. As a result, ERCP, EUS, and MRCP now play a central role in the evaluation and therapy of patients with IAP.

In the past, when EUS and MRCP were not widely available, ERCP was considered a reasonable option to investigate IAP and IARP as it was both diagnostic and offered therapeutic options. There are a number of potential causes of IAP that can be diagnosed and potentially treated by ERCP and ancillary techniques. These include occult gallstones, abnormalities and anomalies of the PD and bile duct, sphincter of Oddi dysfunction (SOD), and ampullary and pancreatic neoplasms. The techniques applied at ERCP to diagnose the cause of IAP are shown in Box 52.2 .

Although there are potential gains in performing ERCP (identifying and treating the cause and preventing another episode of AP), there are potential downsides for the patient and the health care system as a whole (inappropriate performance of the procedure and its adverse events). The timing of ERCP in patients with IAP is controversial. Ballinger and colleagues reported that only 1 of 27 patients with one unexplained episode of pancreatitis with a gallbladder (GB) in situ had a second episode of pancreatitis during a 3-year follow-up period. They felt that the risks of ERCP were greater than the risks of a second episode of AP and advised against its use. On the other hand, Trapnell and Duncan reported that 35 of 148 patients (24%) with IAP suffered a recurrence, but if gallstones were present, the rate increased to 38%. Using an actuarial method, the authors found that 10% of patients with IAP were likely to have a first recurrence within 1 year of the initial attack, 17% within 2 years, and 25% within 6 years. Patients who had one recurrence were likely to have a second. In a cost-utility analysis, Gregor and colleagues found that performing ERCP on all patients after a first episode of IAP was neither of great benefit nor particularly costly. However, it was of substantial benefit and cost-effective in a subgroup of patients with greater probability of having an occult stone.

EUS and MRCP have assumed a central role in the evaluation of patients with IAP and IARP because of their high diagnostic accuracy and low morbidity. The use of secretin in both EUS (secretin-enhanced EUS [S-EUS]) and MRCP (secretin-enhanced MRCP [S-MRCP]) has further improved the diagnostic yield for identifying underlying structural etiologies for IAP. The results of EUS and MRCP can direct an alternative therapy (e.g., cholecystectomy for microlithiasis) and obviate more invasive ERCP. The limitations to these two diagnostic modalities are that they have no therapeutic options and a separate procedure may have to be carried out for treatment. ERCP is now used primarily as a therapeutic option directed by the results of EUS and MRCP, unless the EUS and/or MRCP is negative. In such a setting, sphincter of Oddi manometry (SOM) is usually combined with ERCP.

Our approach to evaluating patients with IAP and IARP has evolved over the years as mounting evidence supported a central role of EUS and MRCP in this setting. Traditionally, for young patients with a negative standard evaluation, we would recommend no further evaluation unless a second episode occurred. In contrast, for patients age 40 years and older in whom findings are negative or inconclusive on standard evaluation, we would then proceed with ERCP (usually with SOM and bile microscopy with a GB in situ if not already done after their first episode of AP). This is based on the findings of Choudari and colleagues. In this study, 21% of patients 40 to 60 years of age and 25% of patients older than 60 years had a neoplastic process as the cause for their pancreatitis, in contrast to only 3% younger than 40 years ( Table 52.1 ).

In a similar series of 1241 patients with idiopathic pancreatitis reported by Fischer and colleagues, the incidence of malignancy and premalignant conditions was 4.7% (58 of 1241). Analysis of a subgroup of patients with intraductal papillary mucinous neoplasm (IPMN), which accounted for 52 of 58 of the malignant and premalignant conditions, showed an increasing trend with age, ranging from 1.3% in patients <40 years to 13% in patients older than 70 years ( Fig. 52.1 ).

Based on these studies, which show an increasing incidence of a neoplastic process with age, and given the critical diagnostic role of EUS and S-MRCP, our current approach is to evaluate patients 40 years and older with EUS and/or S-MRCP after their first episode of IAP to assess for a tumor or other structural cause of the pancreatitis.

The roles of ERCP, EUS, and MRCP in relation to diagnosis and outcomes of therapy for each disease identified will be discussed individually, and a recommended algorithm for the workup of patients with IAP and IARP will be presented.

Occult Gallstone Disease

Although microlithiasis and biliary sludge are technically different, the terms are often used interchangeably. Microlithiasis most commonly refers to stones <3 mm in diameter and biliary sludge is considered to be a suspension of crystals, mucin, glycoproteins, cellular debris, and proteinaceous material within bile. The crystals are made of cholesterol monohydrate, calcium bilirubinate, or calcium carbonate. Microlithiasis and biliary sludge can be found within the GB or bile ducts and might not be detected by standard GB imaging techniques, such as ultrasonography or computed tomography (CT) scan.

Microlithiasis of the GB has been implicated as a common cause of IAP. Two prospective studies found that approximately two-thirds to three-fourths of patients with IAP harbored occult gallstones in the GB. The diagnosis was based on microscopic examination of bile for crystals and usually confirmed on evaluation of the resected GB or follow-up GB ultrasonography showing gallstones and/or sludge. On multivariate analysis, the finding of biliary crystals in bile is a strong predictor of small stones or sludge in the GB ( p < 0.001). Moreover, the finding of crystals in bile had a sensitivity of 86%, specificity of 86%, and positive predictive value of 92% for the diagnosis of gallstone disease as the missed cause of IAP. In contrast to the results of Lee and colleagues and Ros and colleagues, some investigators have detected microlithiasis in less than 10% of patients with IAP. Bile may be collected at the time of ERCP from the duodenum or bile duct after GB stimulation with cholecystokinin or by direct cannulation of the GB.

In patients with IAP, EUS may help identify those with underlying microlithiasis when conventional transcutaneous ultrasonography is normal. Frossard and colleagues reported that EUS identified a biliary cause of IAP in 103 of 168 patients (61%). Of those 103 patients, 52 (50%) had gallstones or microlithiasis, 12 (12%) had GB sludge, 10 (10%) had common bile duct stones, and 29 (28%) had a combination of these findings. Yusoff and colleagues compared the diagnostic yield of EUS in 201 patients with a single episode of IAP with that of 169 patients with recurrent episodes. Biliary tract disease (46 of 246 [18.7%]) was the most common positive finding in the patients with GB in situ. In postcholecystectomy patients, 4 of 124 (3.2%) had evidence of stones in the bile duct. Morris-Stiff and colleagues showed that in 42 patients with idiopathic pancreatitis and a normal MRCP, EUS detected cholelithiasis or microlithiasis alone in 9 (21.4%), cholelithiasis and choledocholithiasis in 6 (14.3%), and choledocholithiasis alone in 1 (2.4%). In another series of 44 patients with IAP, EUS found cholelithiasis in 3 (6.8%), microlithiasis in 20 (45.5%), and choledocholithiasis in 2 (4.5%). Ardengh and colleagues found GB microlithiasis in 27 of 36 patients with IAP (75%). Compared with the final surgical resection specimen, the sensitivity, specificity, and positive and negative predictive values for EUS identification of GB microlithiasis were 92.6% (74.2% to 98.7%), 55.6% (22.7% to 84.7%), 86.2% (67.4% to 95.5%), and 71.4% (30.3% to 94.9%), respectively. Overall EUS accuracy was 83.2%.

Intraductal ultrasonography (IDUS) has also been shown to be useful in detecting occult biliary stones, microlithiasis, and sludge. In a study by Kim and colleagues of 31 patients with IARP and negative findings on ERCP, IDUS revealed small bile duct stones (≤3 mm) in 5 patients (16.1%) and sludge in 3 patients (9.7%).

The role of MRCP in detecting choledocholithiasis is well established. Its role in detecting microlithiasis or sludge of the GB has been extensively studied. Calvo and colleagues evaluated 80 patients undergoing both transabdominal ultrasonography and MRCP for suspected gallstones; the sensitivity of MRCP in diagnosing GB stones (43 patients [97.7%]) was comparable to that of transabdominal ultrasonography (44 patients). MRCP diagnosed biliary sludge or microlithiasis in 13 patients, versus 5 using ultrasonography. The authors concluded that MRCP is a good technique for diagnosing cholelithiasis and biliary sludge. However, high cost, contraindications, and the need for patient cooperation limit MRI for routine GB evaluation.

Treatment of microlithiasis can significantly reduce the incidence of recurrent pancreatitis. There are several therapeutic options for managing patients with pancreatitis caused by microlithiasis. Laparoscopic cholecystectomy should be considered the procedure of choice as it is almost always curative. Ros and colleagues reported no further episodes of pancreatitis in 17 of 18 patients followed for 3 years after cholecystectomy. Endoscopic biliary sphincterotomy is an excellent alternative for elderly and high-risk surgical patients. Dissolution therapy with ursodeoxycholic acid has also been shown to prevent recurrent pancreatitis. However, maintenance therapy is expensive and necessary to prevent recurrent stone formation.

Given the high prevalence of occult microlithiasis in some series, some authors advocate empiric cholecystectomy as a first-line therapy, particularly in patients with recurrent attacks.

Sphincter of Oddi Dysfunction (See Chapter 47 )

SOD refers to the abnormality of sphincter of Oddi contractility that is manifested clinically by pancreaticobiliary pain, pancreatitis, or deranged liver function tests. SOM is considered by most authorities to be the gold-standard test for diagnosing SOD. It is most commonly performed at the time of ERCP but can also be done percutaneously or during surgery. SOM uses a water-perfused catheter that is inserted into the common bile duct, pancreatic duct, or both to measure sphincter pressure. The diagnosis of SOD is established when the basal pressure is ≥40 mm Hg.

Because SOM is difficult to perform, invasive, not widely available, and associated with a high adverse event rate, several noninvasive and provocative tests have been designed in an attempt to identify patients with SOD. The currently available data suggest that these tests lack the sensitivity and specificity to replace SOM.

The role of S-MRCP in diagnosing SOD is still debatable. Mariani and colleagues reported that S-MRCP and SOM were concordant in 13 of 15 patients (86.7%). However, subsequent larger studies did not show this high concordance. Pereira and colleagues showed diagnostic accuracy of S-MRCP for SOD type II and III at 73% and 46%, respectively. Aisen and colleagues also showed that there was no difference in the magnitude of the increase of pancreatic duct diameter between patients with elevated basal sphincter pressure and normal basal sphincter pressure.

Siddiqui et al. reported a low but significant yield of EUS in 143 patients with suspected SOD type 3 with previous normal endoscopies and imaging studies. The role of S-EUS in diagnosing SOD in idiopathic pancreatitis is unclear. The study by Mariani and colleagues labeled 2 patients with SOD because there was persistent main pancreatic duct dilation 15 minutes after secretin injection. There was no recurrent pancreatitis in these patients 18 months after sphincterotomy was performed.

SOD is a frequent cause of IAP. It has been manometrically documented in 15% to 73% of such patients ( Table 52.2 ).

The role of SOM and sphincter ablation for IARP patients is controversial, though a recent survey of American Society for Gastrointestinal Endoscopy members showed considerable variation in approaches to managing patients with SOD and IARP.

Many authorities recommend performing pancreatic SOM in patients with IARP, particularly those with normal biliary manometry and those who have recurrent attacks after a biliary sphincterotomy. Isolated pancreatic sphincter hypertension is commonly found among patients with IARP.

Historically, pancreatic sphincter ablation has been done surgically. However, with the endoscopic sphincterotomy has become the treatment of choice. Studies evaluating the role of SOM and sphincter ablation for patients with IARP began to appear in the late 1980s. The value of ERCP, SOM, and sphincter ablation therapy was studied in 51 patients with IARP. Twenty-four patients (47.1%) had an elevated basal sphincter pressure. Thirty were treated by biliary endoscopic sphincterotomy (BES; n = 20) or surgical sphincteroplasty with septoplasty ( n = 10). Fifteen of 18 patients (83%) with an elevated basal sphincter pressure had long-term benefit (mean follow-up, 38 months) from sphincter ablation therapy (including 10 of 11 treated by BES) in contrast to only 4 of 12 (33.3% p < 0.05) with a normal basal sphincter pressure (including 4 of 9 treated by BES). The results of this study suggested that SOM was predictive of outcome from sphincter ablation in patients with IARP and that BES could prevent recurrent pancreatitis. In the early 1990s, BES became the standard of care for patients with IARP found to have SOD. However, Guelrud and colleagues found that severing the pancreatic sphincter was necessary to resolve pancreatitis ( Fig. 52.2 ).

In this series, 69 patients with IARP caused by SOD underwent treatment by standard BES ( n = 18), BES with pancreatic sphincter balloon dilation ( n = 24), BES followed by pancreatic endoscopic sphincterotomy (PES) in separate sessions ( n = 13), or dual-endoscopic sphincterotomy (DES) ( n = 14). Eighty-one percent of patients undergoing DES had resolution of pancreatitis compared with 28% of patients undergoing BES alone ( p < 0.005). Sherman and colleagues reported that only 44% of SOD patients with IARP had no further attacks during a 5-year follow-up interval after BES. These data are consistent with the theory that many such patients who benefit from BES alone may have subtle gallstone pancreatitis or perhaps follow-up that has not been long enough to detect another attack of pancreatitis. Wehrmann attempted to help clarify the issue by studying patients with a longer follow-up (11.5 ± 1.6 years) after endoscopic therapy for SOD with IARP. In this study, 5 of 37 (14%) had recurrent pancreatitis over a mean duration of 32.4 months (range 24 to 53 months) and this increased to 19 of 37 (51%) at 11.5 years. The frequency of pancreatitis episodes was, however, lower than that before endoscopic therapy. The authors suggest that endoscopic sphincter ablation may slow the progress of the natural course of the disease.

The results of Guelrud and colleagues also support the anatomic findings of separate biliary and pancreatic sphincters, in addition to manometric findings of residual pancreatic sphincter hypertension, in more than 50% of persistently symptomatic patients who undergo BES alone. Kaw and Brodmerkel reported that among patients with idiopathic pancreatitis secondary to SOD, 78% had persistent manometric evidence of pancreatic sphincter hypertension despite biliary sphincterotomy. Toouli and colleagues demonstrated the importance of pancreatic and biliary sphincter ablation in patients with idiopathic recurrent pancreatitis. In this series, 23 of 26 patients (88%) undergoing surgical ablation of both the biliary and pancreatic sphincters were either asymptomatic or had minimal symptoms at a median follow-up of 24 months (range 9 to 105 months). In contrast, only 43% of patients treated conservatively ( n = 7) and 29% treated by BES alone ( n = 7) had no further episodes of pancreatitis. Okolo and colleagues retrospectively evaluated the long-term results of PES in 55 patients with manometrically documented or presumed pancreatic sphincter hypertension (presumption based on recurrent pancreatitis along with PD dilation and contrast medium drainage time from the PD greater than 10 minutes). During a median follow-up of 16 months (range 3 to 52 months), 34 patients (62%) reported significant pain improvement. Patients with normal pancreatography were more likely to respond to therapy than those with pancreatographic evidence of CP (73% vs 58%). Jacob and colleagues postulated that SOD might cause recurrent episodes of pancreatitis even though SOM was normal and that pancreatic stent placement might prevent further attacks. In a randomized study, 34 patients with IARP, normal pancreatic duct SOM, ERCP, secretin testing, and without biliary crystals (probably best considered “true” IARP) were treated with pancreatic stents ( n = 19; 5 to 7 Fr; with stents exchanged 3 times over a 1-year period) or conservative therapy ( n = 15). During a 3-year follow-up, pancreatitis recurred in 53% of patients in the control group and only 11% of the stented patients ( p < 0.02). This study suggests that SOM may be an imperfect test, as patients may have SOD that is not detected at the time of SOM. However, long-term studies are needed to evaluate the outcome after removal of stents, and concerns remain regarding stent-induced ductal and parenchymal changes. Because of the concern of stent-induced injury to the pancreas, trial PD stenting to predict the outcome from PES is not routinely recommended.

Thus the therapeutic landscape had changed. Initial recommendations for BES alone were replaced with dual pancreatic and biliary sphincterotomy for those with SOD. However, no randomized trials were performed with long-term follow-up to support this approach. At present, controversy persists as to the appropriateness of performing SOM in patients with IAP and IARP. In an editorial by Tan and Sherman, it was stated that although the aforementioned studies suggest that endoscopic therapy may benefit a majority of patients with IARP caused by SOD, there are many limitations that need to be emphasized: (1) the majority of published studies are retrospective (except for one study) and suffer from incomplete follow-up, lack homogeneity of patient selection for therapy, and are not blinded or compared with an untreated group; (2) uncontrolled prospective studies are prone to bias; (3) length of follow-up of most studies is less than 3 years; and (4) short follow-up may result in an underestimate of the true recurrence rates.