Gastrointestinal Intervention 2017; 6(1): 32-36
Published online March 31, 2017 https://doi.org/10.18528/gii170005
Copyright © International Journal of Gastrointestinal Intervention.
Rohan M. Modi1, Ravi B. Pavurala1, and Somashekar G. Krishna2,*
1Department of Internal Medicine, The Ohio State University Wexner Medical Center, Columbus, OH, USA, 2Division of Gastroenterology, Hepatology and Nutrition, The Ohio State University Wexner Medical Center, Columbus, OH, USA
Correspondence to: Sections of Pancreatic Disorders and Advanced Endoscopy, Division of Gastroenterology, Hepatology and Nutrition, 395 West 12th Avenue, 2nd floor, Columbus, OH 43210, USA.
This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Pancreatic malignancy is the third leading cause of cancer related death in the United States with limited viable screening options. By the end of this decade, cancers are poised to become the leading cause of death with pancreatic cancer projected to be the second leading cause of cancer related mortality. Pancreatic cystic lesions (PCLs) are found in approximately 5%–14% of patients due to the increased utilization of cross-sectional imaging, with approximately 8%–10% of pancreatic cancers originating as PCLs. Current screening guidelines have shown discrepancies between morphologic characteristics of PCLs and identifying advanced pancreatic disease. Molecular analysis has emerged as a novel technology to aid in adequate diagnosis and management decisions of PCLs. Mucinous cysts including intraductal papillary mucinous neoplasms (IPMNs) or mucinous cystic neoplasms have similar oncogenic mutations including
Keywords: Biological tumor marker, Cystadenoma, serous, Loss of heterozygosity, Neoplasms, cystic, mucinous, and serous, Pancreatic neoplasms
Pancreatic cancer has risen to be the third leading cause of mortality secondary to malignancies in the United States with an estimated mortality of 41,780 people for the year 2016.1 Moreover, the 5-year survival for pancreatic cancer is alarmingly low at 6% and has been unchanged over the past 40 years.2 While we have limited viable screening options for pancreatic cancer, increased utilization of cross-sectional abdominal imaging has resulted in an increase number of pancreatic cystic lesions (PCLs) being discovered. In fact, PCLs are found in approximately 2%–14% of patients who undergo either computed tomography (CT) or magnetic resonance imaging (MRI).3–5 The influx of patients with undifferentiated PCLs has outlined the significance of determining which precursor lesions will progress to pancreatic adenocarcinoma or metastatic disease.6
The current approach for evaluating a PCL is multidisciplinary including radiographic findings, endoscopic ultrasound (EUS), fine needle aspiration (FNA), tumor markers (carcinoembryonic antigen [CEA]), and cytology. The screening approach over the past decade has been outlined by International Consensus Guidelines (2007 Sendai, 2012 Fukuoka), and the American Gastroenterology Association (AGA) (Table 1).7–9 The Sendai guidelines were introduced in 2006 and recommended surgical removal for any suspected MCN, main duct intraductal papillary mucinous neoplasm (MD-IPMN), and mixed duct IPMN.7 Additional criteria for surgery included the following: clinical symptoms, dilated pancreatic duct (≥ 6 mm), intracystic mural nodule, or positive cytology. While these guidelines had a sensitivity of 100%, multiple studies indicated low specificity ranging from 23%–31%.10,11 Subsequently, revised Fukuoka guidelines were presented in 2012 with stricter surgical criteria: pancreatic duct ≥ 10 mm, an enhancing solid component, or obstructive jaundice with a pancreatic head cyst.8 While this resulted in improved specificity, there was a loss in sensitivity to detect advanced neoplasia and no statistical differences when compared to the Sendai guidelines.12
Both Sendai and Fukuoka guidelines are limited as they only focus on IPMN and MCN cysts. The introduction of AGA guidelines in 2015 focused on having a uniform approach for all asymptomatic neoplastic cysts. Compared to Fukuoka, AGA guidelines presented a more rigorous standard for both EUS-FNA and surgical removal that included two high-risk features: size > 3 cm, presence of mural nodule, and dilated main pancreatic duct in the setting of positive cytology.9 The stringent criteria for endoscopic evaluation of PCLs has resulted in concern that following AGA guidelines would result in inadequate identification of progressive neoplastic disease. One retrospective study found that using AGA guidelines on a cohort of 225 patients had lower sensitivity and specificity of 62% and 79% respectively for detecting advanced neoplasia.13 Moreover, the guidelines would have missed 45% of IPMNs with adenocarcinoma or high-grade dysplasia.
Collectively these guidelines indicate that morphologic characteristics are inadequate predictors of advanced pancreatic disease. The introduction of molecular analysis has emerged as a novel technology to assist in the evaluation of PCLs prior to surgical intervention.14,15 The progression from a premalignant cyst to advanced neoplasm can occur through a variety of mechanisms including oncogene mutations, loss of tumor suppression genes, or hypermethylation. While cytologic testing requires optimal levels of pancreatic cyst fluid, the benefit of molecular testing is that minimal aspirate is required from the PCL to assess for genetic alterations. Our aim in this review is to summarize the current literature on distinct gene mutations within subtypes of PCLs for possible integration in diagnostic evaluation of cystic lesions.
The four most common PCLs include intraductal papillary mucinous neoplasm (IPMN), mucinous cystic neoplasm (MCN), serous cystadenoma (SCA), and solid pseudopapillary neoplasm (SPN). IPMNs are most commonly seen in older individuals with the majority located in the pancreatic head or uncinate with expected communication with the main pancreatic duct, while MCNs are typically seen in middle-aged women (40s–50s) primarily in the body or tail of the pancreas. Both IPMNs and MCNs are mucinous cystic lesions and are likely to have elevated CEA in cystic fluid. On the other hand, SPNs are seen in younger women (20s–30s) who present with symptoms with a slight predisposition for the pancreatic body or tail, while SCAs are usually present in older women and are similarly found in the body or tail of the pancras.16
Because the aforementioned characteristics are generalizations, the challenge is an accurate identification of PCL type to adequately guide management decisions. For example, SCAs are benign, non-mucinous lesions that do not require surgical removal or imaging surveillance unless clinically warranted.17,18 On the other hand, SPNs are classically seen in young women and given their aggressive metastatic nature are surgically resected.16,19 Similarly, MCNs can progress to invasive disease and are usually resected rather than continued imaging or endoscopic surveillance.8 Finally, IPMNs are followed due to slow progression from low or intermediate to high or invasive adenocarcinoma.20,21 While low or intermediate grade dysplasia can be monitored through surveillance, more advanced lesions needs to be surgically resected.8 Other rare types of PCLs including cystic pancreatic neuroendocrine tumors (cystic-NETs), squamous lined cysts (lymphoepithelial cysts, epidermoid cysts), and cystic degeneration of metastatic cancers pose additional challenges and diagnosis relies significantly on EUS-FNA cytology.16 The aforementioned management decisions demonstrate the importance of differentiating PCL type with eventual goal of preventing development of pancreatic malignancy.
Recent advancements have allowed for the integration of molecular DNA into the diagnostic approach of PCLs. There are three major components which molecular analysis focuses: (1) DNA quantity and quality, (2) oncogenic mutations (i.e.,
IPMNs are stratified based on the involvement of the main pancreatic duct and include MD-IPMN, branch duct IPMN (BD-IPMN), or mixed IPMN. This classification is important as a recent analysis of the IPMN literature reported malignancy incidence between 57%–92% when the main pancreatic duct is involved (MD-IPMN, mixed IPMN) and 6%–46% in BD-IPMN.22 Furthermore, IPMNs can be characterized on the basis of morphology including gastric, intestinal, pancreatobiliary and oncocytic subtypes.22 Previous molecular studies have focused on DNA quantity and quality,
In terms of oncogenic mutations, both
Several studies have combined DNA quantity,
MCNs are typically seen in middle-aged women and located in the body or tail and are not associated with the main pancreatic duct.16 These cysts are lined by mucin producing tall columnar epithelial cells with mild to severe dysplasia. Based on the International consensus guidelines, MCNs were differentiated from IPMNs based on presence of ovarian stroma.37 Unlike IPMNs, the incidence of
SCAs are benign cystic neoplasms that generally present in older women with a slight predisposition for the body or tail of the pancreas. The neoplasm is a multilobular cyst with a thin capsule that can have radiographic findings such as sunburst pattern of calcification. SCAs consist of single-layer epithelium with glycogen-rich clear cells within cysts.43 Given their indolent slow growth and only 28 cases of serous cystadenocarcinomas in the past 3 decades, the current management of asymptomatic patients is continued surveillance without surgical intervention.18 Molecularly, mutations seen in tumor suppressor gene
SPNs are typically well-demarcated solitary lesions that are found in younger women (mean age of 25 years).19 SPNs have variable features on CT/EUS and can be cystic, solid, or have both cystic and solid components. These lesions do not communicate with the main pancreatic duct and consist of branching papillae with myxoid stroma on cytology.16 Molecular analysis of SPN has fewer genetic alterations compared to other cysts, but always have a missense mutation
Other rare types of PCLs including cystic-NET, squamous lined cysts including lymphoepithelial cysts and epidermoid cysts of accessory spleen, and secondary degeneration of solid primary/ metastatic lesions to the pancreas have not been characterized by molecular studies. Cystic-NETs represent approximately 8% of resected PCLs and are more common than SPTs (4% of resected PCLs).44 A majority of cystic-NETs are indistinguishable from other PCLs on preoperative imaging and cyst-fluid CEA levels are generally very low; however, the yield of EUS-FNA cytology is high (73%) compared to other PCLs (20%).44,45
Cyst fluid molecular markers can play an integral role in not only differentiating the types of PCLs, but also in risk-stratification of mucinous lesions. This facet of research will widen further with the management of PCLs becoming more conservative. Future guidelines will need to decide how to integrate molecular testing into a working algorithm for all PCLs. Recently, the University of Pittsburgh Medical Center based on PanSeq NGS Panel has proposed their own pathway in which there is a lower threshold to perform an EUS-FNA and they have three criteria to consider surgical removal: (1) positive or suspicious cytopathology for malignancy or (2) mucinous cysts > 3 cm with either main pancreatic duct dilation or definitive mural nodule or (3) molecular testing including
Future studies should focus on molecular markers to improve sensitivity of MCNs to ensure adequate management of these PCLs. It is of concern that
PCLs are increasingly detected on cross-sectional imaging studies. The international consensus guidelines indicate that morphologic characteristics are unable to definitely determine advanced neoplasia in cystic lesions. Molecular analysis has emerged as a novel technology to aid in adequate diagnosis and risk stratification of PCLs. Although the diagnostic parameters and techniques for detection of molecular markers are rapidly evolving and increasingly utilized in practice, there is a need for more definitive evidence of their clinical value for wide-spread utilization and definitive incorporation in management guidelines of PCLs.
Criteria for Surgical Removal vs Surveillance in Pancreatic Cystic Lesions Based on Sendai, Fukuoka, and AGA Guidelines
Sendai guideline | Fukuoka guideline | AGA guideline | |
---|---|---|---|
Year | 2006 | 2012 | 2015 |
Criteria for surgical resection or EUS* | Symptomatic | Obstructive jaundice | N/A† |
Size > 3 cm | No size limit | Size > 3 cm | |
Dilated pancreatic duct (≥ 6 mm) | Dilated pancreatic duct (≥ 10 mm) | Dilated pancreatic duct | |
Mural nodule | Enhancing mural nodule | Mural nodule | |
Positive cytology | Positive cytology | Positive cytology | |
Surveillance | < 1 cm: annual MRI vs CT | < 1 cm: MRI vs CT every 2–3 yr | Repeat MRI in 1 yr and then biennially for 5 yr |
1–2 cm: MRI every 6–12 mo | 1–2 cm: annual MRI for a minimum of 2 yr | Positive features on repeat MRI: EUS | |
2–3 cm: MRI every 3–6 mo | 2–3 cm: EUS in 3–6 mo, afterwards alternating MRI/EUS | ||
> 3 cm: alternating MRI/EUS every 3–6 mo | |||
Statistical significance for detection of advanced neoplasia‡ | High sensitivity, low specificity | Lower sensitivity than Sendai | Lower sensitivity and specificity |
AGA, American Gastroenterology Association; EUS, endoscopic ultrasound; N/A, not applicable; MRI, magnetic resonance imaging; CT, computed tomography.
†AGA guidelines only applicable to asymptomatic patients.
‡High grade dysplasia or adenocarcinoma.
Molecular Markers Differentiated by Oncogenic and Tumor Suppressor Genes for IPMN, MCN, SCA, and SPN
Molecular markers | IPMN | MCN | SCA | SPN |
---|---|---|---|---|
Oncogenic | ||||
| + | + | − | − |
| + | − | − | − |
| + | − | − | − |
| + | + | − | + |
| + | + | − | − |
| + | + | − | + |
| + | + | − | − |
| + | + | − | − |
| − | − | − | + |
Tumor suppressor | ||||
| + | − | − | − |
| + | − | − | − |
| − | − | + | − |
© The Society of Gastrointestinal Intervention. Powered by INFOrang Co., Ltd.