Diagnosis and Prognosis of Acute Pancreatitis

EUS is not necessary for the diagnosis of acute pancreatitis (AP), which is usually accomplished by serum enzyme levels, characteristic abdominal pain, or conventional imaging such as computed tomography (CT). The most common EUS feature of AP is hypoechogenicity of the pancreas, indicating edema. Less common findings include pancreatic enlargement and peripancreatic fluid. EUS may also reveal a normal-appearing pancreas in mild cases.

A few studies have explored the role of EUS in predicting the severity of AP. Although imaging with CT and magnetic resonance imaging (MRI) is of greater practical value in this setting, EUS has benefits of no radiation or contrast. One study found that a geographic hyperechoic area within the pancreas was an adverse clinical predictor of severity. Another showed that the presence of peripancreatic edema, parenchymal heterogeneity, dilation of the common bile duct (CBD), and ascites was associated with severe AP in a univariable analysis. These findings require validation before widespread application in the early phase of AP. EUS has a significant role in the management of acute pancreatic pseudocysts and walled-off necrosis, discussed in depth in another section.

Determining Causes of AP

In most cases, the cause of AP is obvious and successfully eliminated, preventing future attacks. For example, if there is radiographic evidence of gallstones, a cholecystectomy usually resolves the issue. Even in these clear-cut cases, it may still be reasonable to check a triglyceride level, screen the medication list, and make sure cancer has been ruled out with appropriate imaging. Beyond these basic considerations, further costly or invasive testing to look for coexisting etiologic factors is not required.

In about 20% of cases of AP, an obvious cause is not found, even after a careful history, ultrasonography, and routine laboratory studies. Although some advocate expectant management after a single episode of unexplained AP, most concur that further evaluation is necessary in those with recurrent idiopathic bouts. This evaluation may include blood tests for metabolic, autoimmune, and genetic causes, and additional structural testing to look for biliary and ductal obstructive disease.

In the past, endoscopic retrograde cholangiopancreatography (ERCP) was the structural test of choice for evaluation of idiopathic recurrent AP (IRAP). Now, EUS and magnetic resonance cholangiopancreatography (MRCP) are advocated as safer options. There are limited studies comparing EUS and MRCP for etiologic diagnosis of AP. In 1 study, EUS and MRCP were performed prospectively in 49 patients with idiopathic AP. The overall yield was significantly higher for EUS compared with MRCP (51% vs 20%, P = .001), mostly because of biliary causes found with EUS and missed with MRCP. Another similar study showed a higher yield overall and for biliary causes in patients with idiopathic AP compared with MRCP and ERCP. Based on these and other studies, EUS may be a preferred test in the evaluation of IRAP, and has a yield considered adequate to justify its cost and minimal risk. Some of the causes of AP detectable with EUS are discussed in the following sections.

Biliary Pancreatitis

Transabdominal ultrasonography (TAUS) is routinely and correctly performed to look for a biliary source for AP. However, AP may result from very small stones missed with TAUS. It may even be inferred that tiny stones smaller than 2 mm (microlithiasis) would be most likely to traverse the cystic duct and impact at the ampullary orifice. EUS provides complete imaging of the gallbladder body, neck, and fundus like TAUS, and superior imaging of the CBD. The finding of echogenic layering sludge in the gallbladder or CBD is strongly suggestive of coexisting microlithiasis and should prompt consideration of cholecystectomy ( Fig. 1 ).

Gallbladder showing layering echogenic sludge and tiny stones ( arrow ).

Several studies have suggested that EUS has incremental value in uncovering a biliary source even after TAUS, CT, or MRCP are negative. In a study of 35 patients with biliary-type pain and negative TAUS, 33 (95%) were found to have sludge or microlithiasis in the gallbladder or CBD on EUS. In 246 patients with idiopathic AP and negative TAUS, EUS revealed sludge or stones in the gallbladder or CBD in 46 (19%) with gallbladder in situ. In 223 patients with suspected mild biliary AP (alanine transaminase level >120 U/L), a definite biliary cause was not found on TAUS, CT, or MRCP in 37 (17%). Thirty-three of these patients underwent an EUS, and stones or sludge in the gallbladder or CBD were found in more than a third.

The yield of the EUS in AP is higher in those with intact gallbladders compared with those who have undergone a cholecystectomy. However, EUS may still identify a biliary source even in a postcholecystectomy patient. EUS carries 94% sensitivity and 95% specificity for detecting CBD stones ( Fig. 2 ). Although MRCP also provides excellent imaging of the biliary tree, EUS is more sensitive for small CBD stones and is better for visualizing the gallbladder.

Head of pancreas showing a shadowing stone in the distal CBD ( arrow ).

EUS is often reserved for IRAP, but it may reveal a biliary cause even in those with a single unexplained attack. In 134 patients with their first attack and intact gallbladders, 27 (21%) were found to have GB or bile duct stones/sludge. Based on these and other data, a recent technical review suggested that EUS should be considered for patients after their first unexplained bout.

Cholecystokinin (CCK)-stimulated collection of bile for crystal analysis may be performed during EUS in those with intact gallbladders. CCK is administered intravenously at a dose of 0.02 μg per kilogram, producing gallbladder contraction and expression of bile into the duodenal lumen within a few minutes. Five or 10 mL of bile is suctioned through the scope into a fluid trap and placed on ice. The specimen is centrifuged and examined under a polarized microscope for detection of cholesterol and bilirubinate crystals. The addition of bile collection increases sensitivity for microlithiasis, and may be routinely performed if EUS imaging does not reveal gallbladder sludge. In 80 patients with single or recurrent AP (RAP) who had no obvious sludge or stones on EUS, bile analysis revealed crystals in 38 (48%). CCK can cause uterine contractions, so women of childbearing age should have a negative pregnancy test immediately before the endoscopy. Crystal analysis has limited reproducibility and clinical implications and as a result is not widely performed.

Occasionally, patients are referred for ERCP while in the throes of biliary AP to detect and extract retained stones. Cautious endoscopists may think twice about this procedure, because most stones pass spontaneously and ERCP may worsen existing AP. ERCP is warranted when signs of cholangitis are present or liver function tests are increasing, but if clinical suspicion for a retained stone is low, a preliminary EUS may help in decision making. If a stone is present, ERCP with extraction can be performed in the same endoscopic session. If no stone is found, the patient can be spared the added risk. This stepwise strategy has been shown to be beneficial, avoiding ERCP in most patients.

Pancreatic Neoplasms

A pancreatic tumor that obstructs the main pancreatic duct (PD) may produce recurrent pancreatitis, likely from increased ductal pressure. In 1 series, AP was a presenting in symptom in 24 of 174 (13.8%) of patients with pancreatic cancer. Pancreatic neuroendocrine tumors may also cause AP. Cancer is of particular concern in patients older than 40 years who develop an unexplained attack, and cross-sectional imaging is recommended if no other obvious cause is found. However, standard cross-sectional imaging tests may miss a pancreatic mass if it is a small lesion, contrast is withheld, or significant inflammation and necrosis masks its presence. EUS is superior for the detection of small pancreatic neoplasms smaller than 2 cm compared with dual-phase CT scan and contrast-enhanced MRI ( Fig. 3 ). EUS also allows tissue sampling through fine-needle aspiration of suspicious hypoechoic areas or masses.

This 53-year-old patient presented with an unexplained bout of AP. The CT was negative except for slight PD prominence. EUS revealed this small mass in the head of the pancreas smaller than 1.5 cm ( arrow ).

Pancreas Divisum

Pancreas divisum is a congenital malunion of the dorsal and ventral PDs and is found in up to 10% of the Western population. It may be a primary cause or a risk modifier in conjunction with genetic or environmental factors in the genesis of RAP. MRCP with or without secretin injection has essentially replaced ERCP in the diagnosis of pancreas divisum. However, EUS may also be helpful in detecting pancreas divisum and evaluating its significance. One study showed superior sensitivity of EUS compared with MRCP (86.7% vs 60%, P <.001) in 45 patients with pancreas divisum confirmed by ERCP.

Several technical criteria have been proposed for radial and linear endosonography, such as the absence of a stack sign (presence of CBD, PD, and portal vein in 1 image), presence of a crossing duct sign (crossed appearance of CBD and PD), or a separate insertion of PDs and bile ducts into the duodenal wall. Existing studies of these criteria are difficult to interpret because of their small sample sizes. The most logical approach may be to trace the PD from its union with the bile duct at the major papilla proximally into the dorsal pancreas. If the PD can be traced, pancreas divisum has been ruled out by definition. One study evaluated this criterion or the visualization of the duct crossing a distinct demarcation of the ventral and dorsal pancreas (dorsal-ventral anlage) in 162 patients using ERCP as the reference standard. The reported sensitivity was 95% (n = 19 with divisum) and specificity was 97%. However, these test performance estimates were calculated only in the 78% of patients in whom adequate endosonographic ductal visualization was obtained. Almost a quarter of the patients were not included because their ductal anatomy could not be adequately assessed, even in expert hands. Based on these results, EUS may be considered useful to rule out divisum if the above maneuver (tracing PD from ampulla to body) can be accomplished. But because many patients have ductal anatomy that cannot be traced adequately with EUS, it may not be as useful to rule in pancreas divisum.

Most patients with pancreas divisum do not experience symptoms. After diagnosing pancreas divisum in a patient with AP, a secondary diagnostic dilemma is to determine its clinical significance. Endoscopic and EUS findings such as a santorinicele and dorsal duct dilation may help clarify if impaired drainage is causing AP, to justify ERCP with minor papillotomy.

Other Structural Causes

Additional structural causes that can be diagnosed with EUS include annular pancreas, choledochocele with anomalous insertion of the PD, and PD strictures.

Also, the endoscopic examination can offer clues. Duodenal or ampullary disease can obstruct pancreatic outflow and produce recurrent bouts of AP. EUS should begin with a careful upper endoscopic examination using a standard upper endoscope. A side-viewing endoscopy can usually be avoided if adequate ampullary views are obtained with the oblique-viewing echoendoscope. Examples of disease that may cause AP diagnosed with endoscopy include upper gastrointestinal Crohn disease, penetrating peptic ulcer, ampullary polyps, ampullary carcinoma, intraluminal windsock diverticulae, and extraluminal periampullary diverticulae. Another significant endoscopic finding is a gaping major papilla with extrusion of mucin, which indicates main-duct intraductal papillary mucinous neoplasm, which may produce AP through mucin plugging. Endoscopy also permits periampullary biopsies, which may assist in the diagnosis of autoimmune pancreatitis.

Sphincter of Oddi Dysfunction

Sphincter of Oddi dysfunction (SOD) and ampullary stenosis may cause RAP, which is a particular consideration after cholecystectomy. ERCP of any kind in patients with suspected sphincter of Oddi manometry (SOM) with or without sphincterotomy comes with a substantial risk of pancreatitis, and efficacy of sphincterotomy is uncertain. Although EUS can neither reliably diagnose nor treat SOD, it does offer a safe and accurate visualization and measurement of the CBD and PD. Proximal enlargement of these ducts may suggest underlying outflow obstruction to support the SOD diagnosis, as in the Milwaukee criteria. EUS with side-viewing endoscopy may help diagnose restenosis after a previous sphincterotomy.

Some investigators have included secretin injection at the time of EUS, with timed measurement of ductal diameters to determine if there is delayed emptying of the duct. One study included ductal measurements every 1 minute after secretin in multiple patients with various conditions producing PD outflow delay. There were 20 patients with suspected SOD who underwent a previous ERCP with SOM as gold standard. In 13 patients with normal SOM, the secretin EUS was normal in 12 (92% specificity). In 7 who had abnormal SOM, the secretin EUS was abnormal in 4 (57% sensitivity). This test is time consuming, requires meticulous performance to ascertain ductal drainage delays, and has not been externally validated to justify its widespread use. There are multiple factors other than the sphincter of Oddi that may affect the results. For example, patients with CP may have less fluid production in response to secretin and a more fibrotic duct wall, and thus less ductal enlargement after secretin.