Distinguishing Pseudocysts from Pancreatic Cystic Neoplasms

Careful exclusion of a pancreatic cystic neoplasm is mandatory before considering draining pseudocysts. Pseudocysts account for at least 75% of all pancreas cysts, but they can be difficult to distinguish from pancreatic cystic neoplasms, retention cysts, and congenital cysts, especially in patients without a clear history of pancreatitis. Retention cysts are nonneoplastic localized areas of dilation of the pancreatic duct owing to a ductal obstruction from an obstructing neoplasm, stricture, or stone. Like pseudocysts, retention cysts are common complications of chronic pancreatitis. Pancreatic cystic neoplasms include those with malignant potential such as mucinous cystic neoplasms, intraductal papillary mucinous neoplasms, solid pseudopapillary neoplasms with cystic components, pancreatic adenocarcinoma, or neuroendocrine tumors with cystic degeneration, and benign serous cystadenoma. Many malignant pancreatic cysts are well treated with surgical resection; however, intervention to drain a cystic neoplasm mistakenly diagnosed as a pseudocyst can compromise any subsequent attempt at operative resection.

A careful clinical history of preceding pancreatitis of a known origin or established chronic pancreatitis can assist in distinguishing between pseudocysts and potential cystic neoplasms. However, this may not be helpful in all cases. Warshaw and colleagues reported on 12 years of experience with pancreatic cystic neoplasms and noted that 37% of lesions had been misdiagnosed as pseudocysts before operation. They found that clinical factors such as abdominal pain and pancreatitis were unhelpful in distinguishing between the two lesions, because both patients with cystic neoplasms and pseudocysts commonly experienced these symptoms. Warshaw and others also found cyst diameter to be a poor distinguishing factor as the mean size of neoplastic cysts is frequently similar to that of pseudocyts. Computed tomography (CT) and magnetic resonance imaging may be helpful in identifying dependent debris within the lesion, suggestive of a pseudocyst, or rim calcification, indicative of neoplasia. External microlobulated morphology and internal septae were more common in neoplasms than in pseudocysts, but neither of these findings attained significance in one study utilizing magnetic resonance imaging. Further, Chalian and colleagues found that CT attenuation was significantly higher in pseudocysts than in mucin-containing cysts (mean of 18.9 HU for pseudocysts, 13.0 HU for mucinous cystic neoplasms, and 11.4 for intraductal papillary mucinous neoplasms), further helping to distinguish these lesions noninvasively.

Endoscopic ultrasonography (EUS) permits visualization of cystic lesions that are within close proximity to the gastrointestinal (GI) lumen and offers the ability for tissue sampling and cyst fluid analysis before pursuing drainage of a potential pseudocyst. Van der Waaij and colleagues analyzed 12 different studies comprising data from 450 patients with pancreatic cyst fluid analysis and found that an amylase level of less than 250 U/L had 98% specificity for serous cystadenoma, mucinous cystadenoma, or mucinous cystadenocarcinoma and virtually excluded the diagnosis of pseudocyst. Carcinoembryonic antigen less than 5 ng/mL showed a specificity of 95% for pseudocyst and serous cystadenoma as did carbohydrate-associate antigen 19 (CA 19-9) concentration less than 37 U/mL with a specificity of 98%. All of these cyst fluid biomarkers showed sensitivities of 50% or less; therefore, these tests are most helpful when the measured value falls below the stated thresholds (ie, the negative predictive value). There are several promising new candidates for cyst fluid biomarkers, such as glycosolation variants of mucins and carcinoembryonic antigen cell adhesion molecules. It would seem that these diagnostic difficulties would favor surgical drainage over percutaneous or endoscopic drainage of cystic lesions, because surgery affords a more generous cyst wall biopsy; however, the magnitude of this benefit is questionable. In one study, the epithelial lining of cystic lesions was partially (5%–98%) missing in 40% to 72% of all major tumor types, implying that even biopsy of a cyst during surgical cystenterostomy can lead to a missed diagnosis of a neoplasm. Currently, no single test can rule out a pancreatic cystic neoplasm; rather, a careful assessment including a detailed clinical history, radiologic studies, and EUS-guided cyst fluid analysis is the most reliable method for distinction.