In the absence of robust prospective data, the conversation about pancreatic cysts has mostly been driven by low-quality evidence, expert consensus and opinion. Hence, when debating an issue such as the role of endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) in pancreatic cysts greater than 3 cm, one can, within reason, construct an argument to support either side of the coin. Having said that, our evolving understanding of the biological behavior of pancreatic cysts has enabled us to draw certain rational conclusions, which form the framework for our management plan for the patient.
The key question in pancreatic cyst management, with the assumption that the cyst is mucinous, remains accurate identification of cysts that harbor malignancy at the time of diagnosis or those that have the potential to develop invasive carcinoma in the future. In other words, identifying patients who have cancer now, and those who will develop cancer in the future, is crucial. While currently, there are no reliable criteria to predict the latter, there are factors — specifically size and cyst characteristics — that can help us assess the former and stratify the risk of malignancy.
When considering the role of EUS-FNA in asymptomatic cysts greater than 3 cm in size, the issue needs to be considered in the context of two basic questions: is cyst size by itself a predictor of cancer and does cross-sectional imaging provide adequate information to risk stratify cysts, hence rendering EUS-FNA redundant?
Although the literature is controversial for cyst size as a predictor of cancer, there is evidence that size matters. In a meta-analysis evaluating the risk of malignancy in patients with intraductal papillary mucinous neoplasms (IPMN), cyst size greater than 3 cm was the feature most strongly associated with malignancy with an OR 62.4 (95 percent CI 30.8-126.3).1
The AGA technical review describes the prevalent risk of malignancy in cysts greater than 3 cm versus less than 3 cm in size as mildly increased with an OR 2.97(95 percent CI 1.82–4.85).2 In the 2006 consensus-based Sendai guidelines, cyst size of greater than 3 cm was one of the triggers for recommending a surgical resection.3
However, in validation studies, these guidelines were found to be flawed, with patients undergoing unnecessary resections for low-risk cysts greater than 3 cm in size. The subsequent Fukuoka and the evidence-based AGA guidelines have advocated a more conservative approach, with the intention of avoiding unnecessary resections.4,5 Nonetheless, even though the import of cyst size has diminished, it has not been relegated to the low-risk category altogether. The Fukoka guidelines use cyst size of greater than 3 cm as a worrisome feature requiring EUS-FNA for risk stratification, while the AGA guideline includes cyst size greater than 3 cm as one of the features that, along with the presence of other high-risk features, should prompt an EUS-FNA. Thus, the available evidence and expert consensus does suggest that patients with cysts greater than 3 cm in size are at a higher risk, and further stratification with EUS-FNA is a reasonable next step.
The subsequent logical question is whether cross-sectional imaging alone can accurately diagnose and risk stratify cysts. The evidence suggests that while we have seen significant improvement over the years in the quality of imaging and reporting, particularly with respect to serous cystadenomas and mainduct IPMNs, both multidetector CT (MDCT) and magnetic resonance imaging (MRI) have suboptimal accuracy in determining the correct histologic diagnosis in other cysts. In a study from a large tertiary-care center, both MDCT and MRI had a limited positive predictive value for classification of small pancreatic cysts into aggressive and nonaggressive categories (MDCT PPV of 25 percent and MRI 50 percent).6 Another study demonstrated an accuracy of only 39 percent for MDCT in predicting malignant potential of pancreatic cystic neoplasms.7 Similarly, Visser, et al., also reported on the relative accuracy of MDCT and MRI in the characterization of pancreatic cysts and found that the reviewer was correct in only 57 percent of cases.8 Even when evaluating specific high-risk stigmata such as mural nodule or solid component, the performance of cross-sectional imaging has been suboptimal.6,9,10 One can reasonably conclude from this data that in a cyst greater than 3 cm in size without any other risk factors, an EUS-FNA should be recommended to improve stratification.
One can reasonably conclude from this data that in a cyst greater than 3 cm in size without any other risk factors, an EUS-FNA should be recommended to improve stratification.
One could question whether EUS-FNA offers incremental advantage over crosssectional imaging studies. EUS, because of its superior resolution, allows for evaluation of mural nodules and associated masses, and the ability to sample the cyst contents for cytological and fluid analysis, and because of these reasons offers an advantage over cross-sectional imaging. This advantage is borne out by data where EUS with or without FNA was significantly superior to MDCT in accurately classifying a cyst as neoplastic (75 percent vs. 48 percent, p less than 0.0001) and in predicting malignancy (49 percent vs. 11 percent, p less than 0.0001). Similarly,
EUS was superior to MRI alone for a correct diagnosis (76 percent vs. 34 percent, p less than 0.0001). Moreover, MRI did not predict malignancy in any patient.11
In another study of 159 patients with incidental cystic lesions, EUS-FNA had an impact on management in 72 percent of cases: 27.6 percent of patients were discharged who were initially recommended for resection, 20.7 percent received surgery, and 6.9 percent received follow ups rather than undergoing resection.12 Even though cyst fluid cytology has pitfalls with an overall accuracy of approximately 50 percent, it is still a
better predictor of malignancy than symptoms, and can detect more cancers in small cysts than high-risk stigmata.13 Finally, in experienced hands, EUS-FNA is a safe and well-tolerated test with an excellent safety profile.
So what should one infer from this argument? First, cysts greater than 3 cm have a small, albeit not zero, risk of harboring a malignancy at the time of diagnosis, and should be managed as such. Second, cross-sectional imaging is not perfect and EUS-FNA should be recommended for risk-stratifying cysts. Third, in the absence of other risk factors, EUS-FNA offers incremental advantage, with cytology being the only available test that will diagnose a malignancy. The role of EUS-FNA in categorizing cysts will only grow as we enter the era of profiling cysts based on mutational analysis.
Dr. Ahmad has no conflicts to disclose.
1. N. Anand, K. Sampath, B.U. Wu, Cyst Features and Risk of Malignancy in Intraductal Papillary Mucinous Neoplasms of the Pancreas: A Meta-Analysis. Clin Gastroenterol Hepatol 2013 August; 11( 8):913–921
2. Scheiman JM, Hwang JH, Moayyedi P. American Gastroenterological Association technical review on the diagnosis and management of pancreatic cysts. Gastroenterology 2015; 148: 824-848
3. M. Tanaka, S. Chari, V. Adsay, et al. International Association of Pancreatology. International consensus guidelines for management of intraductal papillary mucinous neoplasms and mucinous cystic neoplasms of the pancreas. Pancreatology, 6 2006, pp. 17–32
4. M. Tanaka, C. Fernandez-del Castillo, V. Adsay, et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology, 12 2012, pp. 183–197
5. Vege, S.S., Ziring, B., Jain, R., Moayyedi, P. American Gastroenterological Association Institute guideline on the diagnosis and management of asymptomatic neoplastic pancreatic cysts. Gastroenterology 2015, 148 (4), pp. 819-822.
6. Sainani NI, Saokar A, Deshpande V, et al. Comparative performance of MDCT and MRI with MR cholangiopancreatography in characterizing small pancreatic cysts. Am J Roentgenol, 193 2009, pp. 722–731
7. Fisher WE, Hodges SE, Yagnik V, et al. Accuracy of CT in predicting malignant potential of cystic pancreatic neoplasms. HPB (oxford). 2008; 10:483-490
8. Visser B.C, Yeh B.M.,Qayyum A, et al. Characterization of cystic pancreatic masses: relative accuracy of CT and MRI. Am J Roentgenol, 189 2007, pp. 648–656
9. Barron, M.R., Roch, A.M., Waters, J.A. et al. J Gastrointest Surg 2014 18: 447.
10. N. Zhong, L. Zhang, N. Takahashi, et al. Histologic and imaging features of mural nodules in mucinous pancreatic cysts. Clin Gastroenterol Hepatol, 10 2012, pp. 192–198
11. M.A. Khashab, K. Kim, A.M. Lennon, et al. Should we do EUS/FNA on patients with pancreatic cysts? The incremental diagnostic yield of EUS over CT/MRI for prediction of cystic neoplasms. Pancreas, 42 2013, pp. 717–721
12. Ardengh JC, Lopes CV, de Lima-Filho ER, Kemp R, Dos Santos JS. Impact of endoscopic ultrasoundguided fine-needle aspiration on incidental pancreatic cysts: a prospective study. Scand J Gastroenterol. 2014;49(1):114-120
13. M. Genevay, M. Mino-Kenudson, K. Yaeger, et al. Cytology adds value to imaging studies for risk assessment of malignancy in pancreatic mucinous cysts. Ann Surg, 254 2011, pp. 977–983