Pancreatic cystic lesions are widely identified now because of the increased use and improved resolution of cross-sectional imaging. The majority of these cysts are found in older patients, often when imaging is performed for an unrelated condition such as our patient with a suspected kidney stone. This leads to tremendous anxiety for the patient and provider to characterize the etiology of the cyst, and most importantly, its malignant potential.
While there are small numbers of very uncommon cyst types in the pancreas, most cysts are either non-neoplastic or benign without malignant potential, such as pseudocysts, serous cystadenomas, or lesions that are neoplasms with some risk for malignancy. These include mucinous cystic neoplasms (MCN), and intraductal papillary mucinous neoplasms (IPMN). Historical literature has suggested that pseudocysts were the most common cysts, but it is clear now with improved imaging that IPMNs represent the majority of these incidental lesions.
AGA recently published the first evidence-based guidelines to guide clinicians facing this common clinical problem.1 Developed through a multistakeholder process utilizing the GRADE method to assess the quality of evidence from a comprehensive technical review,2 it offers guidance on all incidentally found pancreatic cysts, recognizing that invasive testing with endoscopic ultrasound (EUS) is often performed in an attempt to characterize the cyst, often with misclassification and perhaps leading to unnecessary surgery.
The key decision points for a pancreatic cystic lesion are 1) is the cyst neoplastic or not? and 2) if neoplastic, what is the risk of malignancy? In an asymptomatic patient, such as the case presented, we cannot make them better, we can only alter the natural history of a condition, which can be in a positive or a negative way, because there is harm both with and without further testing in some cases, and the same thing can be said for treatment.
Cross-sectional imaging can be diagnostic for the pancreatic cyst type. There are classic morphologic features of microcystic serous cystadenomas, including a central scar with calcification. However not every serous cystadenoma has this appearance, and although they are far more common in women, men can have them as well. Thus, a macrocystic serous cystadenoma can engender diagnostic uncertainty, and in selected circumstances, one could proceed with further invasive testing to attempt to confirm a diagnosis to obviate for surveillance. IPMN in the pancreas is by far the most common lesion and is distinguished from an MCN by ductal communication. As our patient is a man, MCN is quite unlikely, with IPMN vs. a nonneoplastic cyst the primary diagnostic differential. While high-quality magnetic resonance imaging (MRI) with MRCP can often make this distinction, as can EUS in highly skilled hands, the tests are neither 100 percent sensitive, nor specific. Thus, the approach to all cysts must consider these diagnostic and therapeutic uncertainties.
The AGA guideline recommends that patients with a pancreatic cyst under 3 cm without a solid component or a dilated duct undergo surveillance. MRI is preferred because of its non-invasive nature and lack of radiation, and, if the cyst is stable after the second imaging test, that it can be followed every two years, for a total of five years. Viewed by some as very controversial recommendations, recall that the vast majority of these cystic lesions are seen in older patients. Thus, the potential of the cyst to cause harm has to be balanced with the patient’s risk of pancreatic resection, both immediate and long-term, as well as the likelihood that the patient’s life expectancy, as well as future quality of life, will be maximized.
Case Example: A 65-year-old male with back pain undergoes a CT colonography and an incidental 3 cm pancreatic cyst is found in the pancreatic head. There is no mass or ductal dilatation. There is also no history or risk factors for pancreatic disease.
We performed a systematic review of the prevalence of cysts in the general population and used the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) database to estimate mortality from pancreatic adenocarcinoma. The overall prevalence of cysts was 15 percent; the prevalence of cysts greater than 2 cm was just under 1 percent. There was a dramatic increasing prevalence with age, such that 25 percent of patients in the seventh decade of life were found to have a cyst, and the cyst prevalence great than 2 cm approached 2 percent in that patient population. We made estimates of the point prevalence of malignancy using SEER data on cystic adenocarcinomas, as well as pancreatic adenocarcinoma, and both were vanishingly small, all less than 1 percent. In the review, pancreatic cysts larger than 3 cm increased risk nearly three-fold. While this is a significant increase in risk, due to low baseline absolute risk, an asymptomatic cyst without worrisome features has extremely low risk of malignancy — even at 3 cm in size.
EUS cannot be routinely justified for all pancreatic cysts. There is little doubt that EUS is the best test for identifying small masses and associated nodules. But it has real challenges in differentiating a mural nodule from associated mucus and is markedly operator dependent (as are all imaging tests). Despite considerable research in the past decades, currently the only clinically useful tumor marker to differentiate a mucinous from non-mucinous lesion is the cyst fluid carcinoembryonic antigen (CEA). We know that a carcinoembryonic antigen level greater than 192 was first demonstrated to differentiate between a mucinous from a non-mucinous lesion with an AUC of .79. This unfortunately means that as many as one in five patients will be misclassified. A cyst without worrisome features on cross sectional imaging, such as the patient presented today, rarely will have
worrisome features on EUS.
The data that we have regarding the outcome of surveillance with imaging evaluated 6,000 patient-years of follow up noted only 42 invasive cancers — a rate of malignancy development of .24 percent per year. This is quite analogous to our evolution in the understanding of other pre-cancerous conditions. Barrett’s esophagus is a great example, where we initially thought all dysplasia became cancer, and that the risk of malignancy was extraordinary. We know now that the rate is well less than 1 percent and that very aggressive intervention in that condition, which carries far less morbidity and mortality for surgery than pancreatic surgery, should be
balanced with a very low risk.
In summary, the AGA guideline informs clinicians across specialties that invasive cancer in asymptomatic pancreatic cysts is actually quite rare. We have to be aware that we are applying diagnostic tests such EUS-FNA
of limited sensitivity and specificity to low disease prevalence populations that rarely alter post-test probability. The fear of “missing cancer” may cause a large number of patients to undergo unnecessary surgery; intensive surveillance may actually lead to the same undesirable outcome due to false positives.
The AGA guideline with its guiding principal that less-intense follow up and less surgery is justified seems appropriate to our patient and strongly argues for a conservative imaging approach.
Dr. Scheiman provided advice regarding study design and clinical application to Teva, Aralez/Pozen, Pfizer and Stryker.
1. Vege SS, Ziring B, Jain R, Moayyedi P. American Gastroenterological Association Institute Guideline on the Diagnosis and Management of Asymptomatic Neoplastic Pancreatic Cysts. Gastroenterology 2015; 148: 819-22
2. Scheiman JM, Hwang JH, Moayyedi P. American Gastroenterological Association Technical Review on the Diagnosis and Management of Asymptomatic Neoplastic Pancreatic Cysts. Gastroenterology 2015; 148: 824-48