User login
Dear colleagues,
Pancreas cysts have become almost ubiquitous in this era of high-resolution cross-sectional imaging. They are a common GI consult with patients and providers worried about the potential risk of malignant transformation. Despite significant research over the past few decades, predicting the natural history of these cysts, especially the side-branch intraductal papillary mucinous neoplasms (IPMNs), remains difficult. There have been a variety of expert recommendations and guidelines, but heterogeneity exists in management especially regarding timing of endoscopic ultrasound, imaging surveillance, and cessation of surveillance. Some centers will present these cysts at multidisciplinary conferences, while others will follow general or local algorithms. In this issue of Perspectives, Dr. Lauren G. Khanna, assistant professor of medicine at NYU Langone Health, New York, and Dr. Santhi Vege, professor of medicine at the Mayo Clinic, Rochester, Minn., present updated and differing approaches to managing these cysts. Which side of the debate are you on? We welcome your thoughts, questions and input– share with us on Twitter @AGA_GIHN
Gyanprakash A. Ketwaroo, MD, MSc, is associate professor of medicine, Yale University, New Haven, Conn., and chief of endoscopy at West Haven (Conn.) VA Medical Center. He is an associate editor for GI & Hepatology News.
Continuing pancreas cyst surveillance indefinitely is reasonable
BY LAUREN G. KHANNA, MD, MS
Pancreas cysts remain a clinical challenge. The true incidence of pancreas cysts is unknown, but from MRI and autopsy series, may be up to 50%. Patients presenting with a pancreas cyst often have significant anxiety about their risk of pancreas cancer. We as a medical community initially did too; but over the past few decades as we have gathered more data, we have become more comfortable observing many pancreas cysts. Yet our recommendations for how, how often, and for how long to evaluate pancreas cysts are still very much under debate; there are multiple guidelines with discordant recommendations. In this article, I will discuss my approach to patients with a pancreas cyst.
At the first evaluation, I review available imaging to see if there are characteristic features to determine the type of pancreas cyst: IPMN (including main duct, branch duct, or mixed type), serous cystic neoplasm (SCA), mucinous cystic neoplasm (MCN), solid pseudopapillary neoplasm (SPN), cystic neuroendocrine tumor (NET), or pseudocyst. I also review symptoms, including abdominal pain, weight loss, history of pancreatitis, and onset of diabetes, and check hemoglobin A1c and Ca19-9. I often recommend magnetic resonance cholangiopancreatography (MRCP) if it has not already been obtained and is feasible (that is, if a patient does not have severe claustrophobia or a medical device incompatible with MRI). If a patient is not a candidate for treatment should a pancreatic malignancy be identified, because of age, comorbidities, or preference, I recommend no further evaluation.
Where cyst type remains unclear despite MRCP, and for cysts over 2 cm, I recommend endoscopic ultrasound (EUS) for fluid sampling to assist in determining cyst type and to rule out any other high-risk features. In accordance with international guidelines, if a patient has any concerning imaging features, including main pancreatic duct dilation >5 mm, solid component or mural nodule, or thickened or enhancing duct walls, regardless of cyst size, I recommend EUS to assess for and biopsy any solid component and to sample cyst fluid to examine for dysplasia. Given the lower sensitivity of CT for high-risk features, if MRCP is not feasible, for cysts 1-2 cm, I recommend EUS for better evaluation.
If a cyst is determined to be a cystic NET; main duct or mixed-type IPMN; MCN; or SPN; or a branch duct IPMN with mural nodule, high-grade dysplasia, or adenocarcinoma, and the patient is a surgical candidate, I refer the patient for surgical evaluation. If a cyst is determined to be an SCA, the malignant potential is minimal, and patients do not require follow-up. Patients with a pseudocyst are managed according to their clinical scenario.
Many patients have a proven or suspected branch duct IPMN, an indeterminate cyst, or multiple cysts. Cyst management during surveillance is then determined by the size of the largest cyst and stability of the cyst(s). Of note, patients with an IPMN also have been shown to have an elevated risk of concurrent pancreas adenocarcinoma, which I believe is one of the strongest arguments for heightened surveillance of the entire pancreas in pancreas cyst patients. EUS in particular can identify small or subtle lesions that are not detected by cross-sectional imaging.
If a patient has no prior imaging, in accordance with international and European guidelines, I recommend the first surveillance MRCP at a 6-month interval for cysts <2 cm, which may offer the opportunity to identify rapidly progressing cysts. If a patient has previous imaging available demonstrating stability, I recommend surveillance on an annual basis for cysts <2 cm. For patients with a cyst >2 cm, as above, I recommend EUS, and if there are no concerning features on imaging or EUS, I then recommend annual surveillance.
While the patient is under surveillance, if there is more than minimal cyst growth, a change in cyst appearance, or development of any imaging high-risk feature, pancreatitis, new onset or worsening diabetes, or elevation of Ca19-9, I recommend EUS for further evaluation and consideration of surgery based on EUS findings. If an asymptomatic cyst <2 cm remains stable for 5 years, I offer patients the option to extend imaging to every 2 years, if they are comfortable. In my experience, though, many patients prefer to continue annual imaging. The American Gastroenterological Association guidelines promote stopping surveillance after 5 years of stability, however there are studies demonstrating development of malignancy in cysts that were initially stable over the first 5 years of surveillance. Therefore, I discuss with patients that it is reasonable to continue cyst surveillance indefinitely, until they would no longer be interested in pursuing treatment of any kind if a malignant lesion were to be identified.
There are two special groups of pancreas cyst patients who warrant specific attention. Patients who are at elevated risk of pancreas adenocarcinoma because of an associated genetic mutation or a family history of pancreatic cancer already may be undergoing annual pancreas cancer screening with either MRCP, EUS, or alternating MRCP and EUS. When these high-risk patients also have pancreas cysts, I utilize whichever strategy would image their pancreas most frequently and do not extend beyond 1-year intervals. Another special group is patients who have undergone partial pancreatectomy for IPMN. As discussed above, given the elevated risk of concurrent pancreas adenocarcinoma in IPMN patients, I recommend indefinite continued surveillance of the remaining pancreas parenchyma in these patients.
Given the prevalence of pancreas cysts, it certainly would be convenient if guidelines were straightforward enough for primary care physicians to manage pancreas cyst surveillance, as they do for breast cancer screening. However, the complexities of pancreas cysts necessitate the expertise of gastroenterologists and pancreas surgeons, and a multidisciplinary team approach is best where possible.
Dr. Khanna is chief, advanced endoscopy, Tisch Hospital; director, NYU Advanced Endoscopy Fellowship; assistant professor of medicine, NYU Langone Health. Email: Lauren.Khanna@nyulangone.org. There are no relevant conflicts to disclose.
References
Tanaka M et al. Pancreatology. 2017 Sep-Oct;17(5):738-75.
Sahora K et al. Eur J Surg Oncol. 2016 Feb;42(2):197-204.
Del Chiaro M et al. Gut. 2018 May;67(5):789-804
Vege SS et al. Gastroenterology. 2015 Apr;148(4):819-22
Petrone MC et al. Clin Transl Gastroenterol. 2018 Jun 13;9(6):158
Pancreas cysts: More is not necessarily better!
BY SANTHI SWAROOP VEGE, MD
Pancreas cysts (PC) are very common, incidental findings on cross-sectional imaging, performed for non–pancreas-related symptoms. The important issues in management of patients with PC in my practice are the prevalence, natural history, frequency of occurrence of high-grade dysplasia (HGD) and/or pancreatic cancer (PDAC), concerning clinical symptoms and imaging findings, indications for EUS and fine-needle aspiration cytology, ideal method and frequency of surveillance, indications for surgery (up front and during follow-up), follow-up after surgery, stopping surveillance, costs, and unintentional harms of management. Good population-based evidence regarding many of the issues described above does not exist, and all information is from selected clinic, radiology, EUS, and surgical cohorts (very important when trying to assess the publications). Cohort studies should start with all PC undergoing surveillance and assess various outcomes, rather than looking backward from EUS or surgical cohorts.
The 2015 American Gastroenterological Association guidelines on asymptomatic neoplastic pancreas cysts, which I coauthored, recommend, consistent with principles of High Value Care (minimal unintentional harms and cost effectiveness), that two of three high-risk features (mural nodule, cyst size greater than 3 cm, and dilated pancreatic duct) be present for EUS-guided fine-needle aspiration (EUS-FNA). By the same token, they advise surgery for those with two of three high-risk features and or concerning features on EUS and cytology. Finally, they suggest stopping surveillance at 5 years if there are no significant changes. Rigorous GRADE methodology along with systematic review of all relevant questions (rather than cohorts of 500 or fewer patients) formed the basis of the guidelines. Those meta-analyses showed that risk of PDAC in mural nodules, cyst size >3 cm, and dilated pancreatic duct, while elevated, still is very low in absolute terms. Less than 20% of resections for highly selected, high-risk cysts showed PDAC. The guidelines were met with a lot of resistance from several societies and physician groups. The recommendations for stopping surveillance after 5 years and no surveillance for absent or low-grade dysplasia after surgery are hotly contested, and these areas need larger, long-term studies.
The whole area of cyst fluid molecular markers that would suggest mucinous type (KRAS and GNAS mutations) and, more importantly, the presence or imminent development of PDAC (next-generation sequencing or NGS) is an exciting field. One sincerely hopes that there will be a breakthrough in this area to achieve the holy grail. Cost effectiveness studies demonstrate the futility of existing guidelines and favor a less intensive approach. Guidelines are only a general framework, and management of individual patients in the clinic is entirely at the discretion of the treating physician. One should make every attempt to detect advanced lesions in PC, but such effort should not subject a large majority of patients to unintentional harms by overtreatment and add further to the burgeoning health care costs in the country.
PC are extremely common (10% of all abdominal imaging), increase with age, are seen in as many as 40%-50% of MRI examinations for nonpancreatic indications, and most (>50%) are IPMNs. Most of the debate centers around the concerns of PDAC and/or HGD associated with mucinous cysts (MCN, IPMN, side-branch, main duct, or mixed).
The various guidelines by multiple societies differ in some aspects, such as in selection of patients based on clinical, laboratory, and imaging findings for up-front surgery or surveillance, the frequency of surveillance based on the size of the cyst and the presence of other concerning cyst features (usually with MRCP), the indications for EUS (both initial and subsequent), importance of the magnitude of growth (most IPMNs slowly grow over a period of time), indications for surgery during surveillance and postsurgery surveillance, and the decision to stop surveillance at some point in time. The literature is replete with small case series reporting a proportion of cancers detected and often ignoring the harms of surgery. Incidence of and mortality caused by PDAC are very low (about 1% for both) in a large national cohort of VA pancreatic cyst patients with long-term follow-up and other studies.
Marcov modeling suggests that none of the guidelines would lead to cost-effective care with low mortality because of overtreatment of low-risk lesions, and a specificity of 67% or more for PDAC/HGB is required. AGA guidelines came close to it but with low sensitivity. Monte Carlo modeling suggests that less intensive strategies, compared with more intensive, result in a similar number of deaths at a much lower cost. While molecular markers in PC fluid are reported to increase the specificity of PDAC/HGD to greater than 70%, it should be observed that such validation was done in a small percentage of patients who had both those markers and resection.
The costs of expensive procedures like EUS, MRI, and surgery, the 3% complication rate with EUS-FNA (primarily acute pancreatitis), and the 1% mortality and approximately 20%-30% morbidity with surgery (bleeding, infection, fistula) and postpancreatectomy diabetes of approximately 30% in the long run need special attention.
In conclusion, one could say pancreas cysts are extremely frequent, most of the neoplastic cysts are mucinous (IPMN and MCN) and slowly growing over time without an associated cancer, and the greatest need at this time is to identify the small proportion of such cysts with PDAC and/or HGD. Until such time, judicious selection of patients for surveillance and reasonable intervals of such surveillance with selective use of EUS will help identify patients requiring resection. In our enthusiasm to detect every possible pancreatic cancer, we should not ignore the unintentional outcomes of surgery to a large majority of patients who would never develop PDAC and the astronomical costs associated with such practice.
Dr. Vege is professor of medicine at the Mayo Clinic. He reported having no conflicts of interest regarding this article.
References
Vege SS et al. Gastroenterology. 2015;148:819-22.
Lobo JM et al. Surgery. 2020;168:601-9.
Lennon AM and Vege SS. Clin Gastroenterol Hepatol. 2022;20:1663-7.
Harris RP. Ann Intern Med. 2015;162:787-9.
Dear colleagues,
Pancreas cysts have become almost ubiquitous in this era of high-resolution cross-sectional imaging. They are a common GI consult with patients and providers worried about the potential risk of malignant transformation. Despite significant research over the past few decades, predicting the natural history of these cysts, especially the side-branch intraductal papillary mucinous neoplasms (IPMNs), remains difficult. There have been a variety of expert recommendations and guidelines, but heterogeneity exists in management especially regarding timing of endoscopic ultrasound, imaging surveillance, and cessation of surveillance. Some centers will present these cysts at multidisciplinary conferences, while others will follow general or local algorithms. In this issue of Perspectives, Dr. Lauren G. Khanna, assistant professor of medicine at NYU Langone Health, New York, and Dr. Santhi Vege, professor of medicine at the Mayo Clinic, Rochester, Minn., present updated and differing approaches to managing these cysts. Which side of the debate are you on? We welcome your thoughts, questions and input– share with us on Twitter @AGA_GIHN
Gyanprakash A. Ketwaroo, MD, MSc, is associate professor of medicine, Yale University, New Haven, Conn., and chief of endoscopy at West Haven (Conn.) VA Medical Center. He is an associate editor for GI & Hepatology News.
Continuing pancreas cyst surveillance indefinitely is reasonable
BY LAUREN G. KHANNA, MD, MS
Pancreas cysts remain a clinical challenge. The true incidence of pancreas cysts is unknown, but from MRI and autopsy series, may be up to 50%. Patients presenting with a pancreas cyst often have significant anxiety about their risk of pancreas cancer. We as a medical community initially did too; but over the past few decades as we have gathered more data, we have become more comfortable observing many pancreas cysts. Yet our recommendations for how, how often, and for how long to evaluate pancreas cysts are still very much under debate; there are multiple guidelines with discordant recommendations. In this article, I will discuss my approach to patients with a pancreas cyst.
At the first evaluation, I review available imaging to see if there are characteristic features to determine the type of pancreas cyst: IPMN (including main duct, branch duct, or mixed type), serous cystic neoplasm (SCA), mucinous cystic neoplasm (MCN), solid pseudopapillary neoplasm (SPN), cystic neuroendocrine tumor (NET), or pseudocyst. I also review symptoms, including abdominal pain, weight loss, history of pancreatitis, and onset of diabetes, and check hemoglobin A1c and Ca19-9. I often recommend magnetic resonance cholangiopancreatography (MRCP) if it has not already been obtained and is feasible (that is, if a patient does not have severe claustrophobia or a medical device incompatible with MRI). If a patient is not a candidate for treatment should a pancreatic malignancy be identified, because of age, comorbidities, or preference, I recommend no further evaluation.
Where cyst type remains unclear despite MRCP, and for cysts over 2 cm, I recommend endoscopic ultrasound (EUS) for fluid sampling to assist in determining cyst type and to rule out any other high-risk features. In accordance with international guidelines, if a patient has any concerning imaging features, including main pancreatic duct dilation >5 mm, solid component or mural nodule, or thickened or enhancing duct walls, regardless of cyst size, I recommend EUS to assess for and biopsy any solid component and to sample cyst fluid to examine for dysplasia. Given the lower sensitivity of CT for high-risk features, if MRCP is not feasible, for cysts 1-2 cm, I recommend EUS for better evaluation.
If a cyst is determined to be a cystic NET; main duct or mixed-type IPMN; MCN; or SPN; or a branch duct IPMN with mural nodule, high-grade dysplasia, or adenocarcinoma, and the patient is a surgical candidate, I refer the patient for surgical evaluation. If a cyst is determined to be an SCA, the malignant potential is minimal, and patients do not require follow-up. Patients with a pseudocyst are managed according to their clinical scenario.
Many patients have a proven or suspected branch duct IPMN, an indeterminate cyst, or multiple cysts. Cyst management during surveillance is then determined by the size of the largest cyst and stability of the cyst(s). Of note, patients with an IPMN also have been shown to have an elevated risk of concurrent pancreas adenocarcinoma, which I believe is one of the strongest arguments for heightened surveillance of the entire pancreas in pancreas cyst patients. EUS in particular can identify small or subtle lesions that are not detected by cross-sectional imaging.
If a patient has no prior imaging, in accordance with international and European guidelines, I recommend the first surveillance MRCP at a 6-month interval for cysts <2 cm, which may offer the opportunity to identify rapidly progressing cysts. If a patient has previous imaging available demonstrating stability, I recommend surveillance on an annual basis for cysts <2 cm. For patients with a cyst >2 cm, as above, I recommend EUS, and if there are no concerning features on imaging or EUS, I then recommend annual surveillance.
While the patient is under surveillance, if there is more than minimal cyst growth, a change in cyst appearance, or development of any imaging high-risk feature, pancreatitis, new onset or worsening diabetes, or elevation of Ca19-9, I recommend EUS for further evaluation and consideration of surgery based on EUS findings. If an asymptomatic cyst <2 cm remains stable for 5 years, I offer patients the option to extend imaging to every 2 years, if they are comfortable. In my experience, though, many patients prefer to continue annual imaging. The American Gastroenterological Association guidelines promote stopping surveillance after 5 years of stability, however there are studies demonstrating development of malignancy in cysts that were initially stable over the first 5 years of surveillance. Therefore, I discuss with patients that it is reasonable to continue cyst surveillance indefinitely, until they would no longer be interested in pursuing treatment of any kind if a malignant lesion were to be identified.
There are two special groups of pancreas cyst patients who warrant specific attention. Patients who are at elevated risk of pancreas adenocarcinoma because of an associated genetic mutation or a family history of pancreatic cancer already may be undergoing annual pancreas cancer screening with either MRCP, EUS, or alternating MRCP and EUS. When these high-risk patients also have pancreas cysts, I utilize whichever strategy would image their pancreas most frequently and do not extend beyond 1-year intervals. Another special group is patients who have undergone partial pancreatectomy for IPMN. As discussed above, given the elevated risk of concurrent pancreas adenocarcinoma in IPMN patients, I recommend indefinite continued surveillance of the remaining pancreas parenchyma in these patients.
Given the prevalence of pancreas cysts, it certainly would be convenient if guidelines were straightforward enough for primary care physicians to manage pancreas cyst surveillance, as they do for breast cancer screening. However, the complexities of pancreas cysts necessitate the expertise of gastroenterologists and pancreas surgeons, and a multidisciplinary team approach is best where possible.
Dr. Khanna is chief, advanced endoscopy, Tisch Hospital; director, NYU Advanced Endoscopy Fellowship; assistant professor of medicine, NYU Langone Health. Email: Lauren.Khanna@nyulangone.org. There are no relevant conflicts to disclose.
References
Tanaka M et al. Pancreatology. 2017 Sep-Oct;17(5):738-75.
Sahora K et al. Eur J Surg Oncol. 2016 Feb;42(2):197-204.
Del Chiaro M et al. Gut. 2018 May;67(5):789-804
Vege SS et al. Gastroenterology. 2015 Apr;148(4):819-22
Petrone MC et al. Clin Transl Gastroenterol. 2018 Jun 13;9(6):158
Pancreas cysts: More is not necessarily better!
BY SANTHI SWAROOP VEGE, MD
Pancreas cysts (PC) are very common, incidental findings on cross-sectional imaging, performed for non–pancreas-related symptoms. The important issues in management of patients with PC in my practice are the prevalence, natural history, frequency of occurrence of high-grade dysplasia (HGD) and/or pancreatic cancer (PDAC), concerning clinical symptoms and imaging findings, indications for EUS and fine-needle aspiration cytology, ideal method and frequency of surveillance, indications for surgery (up front and during follow-up), follow-up after surgery, stopping surveillance, costs, and unintentional harms of management. Good population-based evidence regarding many of the issues described above does not exist, and all information is from selected clinic, radiology, EUS, and surgical cohorts (very important when trying to assess the publications). Cohort studies should start with all PC undergoing surveillance and assess various outcomes, rather than looking backward from EUS or surgical cohorts.
The 2015 American Gastroenterological Association guidelines on asymptomatic neoplastic pancreas cysts, which I coauthored, recommend, consistent with principles of High Value Care (minimal unintentional harms and cost effectiveness), that two of three high-risk features (mural nodule, cyst size greater than 3 cm, and dilated pancreatic duct) be present for EUS-guided fine-needle aspiration (EUS-FNA). By the same token, they advise surgery for those with two of three high-risk features and or concerning features on EUS and cytology. Finally, they suggest stopping surveillance at 5 years if there are no significant changes. Rigorous GRADE methodology along with systematic review of all relevant questions (rather than cohorts of 500 or fewer patients) formed the basis of the guidelines. Those meta-analyses showed that risk of PDAC in mural nodules, cyst size >3 cm, and dilated pancreatic duct, while elevated, still is very low in absolute terms. Less than 20% of resections for highly selected, high-risk cysts showed PDAC. The guidelines were met with a lot of resistance from several societies and physician groups. The recommendations for stopping surveillance after 5 years and no surveillance for absent or low-grade dysplasia after surgery are hotly contested, and these areas need larger, long-term studies.
The whole area of cyst fluid molecular markers that would suggest mucinous type (KRAS and GNAS mutations) and, more importantly, the presence or imminent development of PDAC (next-generation sequencing or NGS) is an exciting field. One sincerely hopes that there will be a breakthrough in this area to achieve the holy grail. Cost effectiveness studies demonstrate the futility of existing guidelines and favor a less intensive approach. Guidelines are only a general framework, and management of individual patients in the clinic is entirely at the discretion of the treating physician. One should make every attempt to detect advanced lesions in PC, but such effort should not subject a large majority of patients to unintentional harms by overtreatment and add further to the burgeoning health care costs in the country.
PC are extremely common (10% of all abdominal imaging), increase with age, are seen in as many as 40%-50% of MRI examinations for nonpancreatic indications, and most (>50%) are IPMNs. Most of the debate centers around the concerns of PDAC and/or HGD associated with mucinous cysts (MCN, IPMN, side-branch, main duct, or mixed).
The various guidelines by multiple societies differ in some aspects, such as in selection of patients based on clinical, laboratory, and imaging findings for up-front surgery or surveillance, the frequency of surveillance based on the size of the cyst and the presence of other concerning cyst features (usually with MRCP), the indications for EUS (both initial and subsequent), importance of the magnitude of growth (most IPMNs slowly grow over a period of time), indications for surgery during surveillance and postsurgery surveillance, and the decision to stop surveillance at some point in time. The literature is replete with small case series reporting a proportion of cancers detected and often ignoring the harms of surgery. Incidence of and mortality caused by PDAC are very low (about 1% for both) in a large national cohort of VA pancreatic cyst patients with long-term follow-up and other studies.
Marcov modeling suggests that none of the guidelines would lead to cost-effective care with low mortality because of overtreatment of low-risk lesions, and a specificity of 67% or more for PDAC/HGB is required. AGA guidelines came close to it but with low sensitivity. Monte Carlo modeling suggests that less intensive strategies, compared with more intensive, result in a similar number of deaths at a much lower cost. While molecular markers in PC fluid are reported to increase the specificity of PDAC/HGD to greater than 70%, it should be observed that such validation was done in a small percentage of patients who had both those markers and resection.
The costs of expensive procedures like EUS, MRI, and surgery, the 3% complication rate with EUS-FNA (primarily acute pancreatitis), and the 1% mortality and approximately 20%-30% morbidity with surgery (bleeding, infection, fistula) and postpancreatectomy diabetes of approximately 30% in the long run need special attention.
In conclusion, one could say pancreas cysts are extremely frequent, most of the neoplastic cysts are mucinous (IPMN and MCN) and slowly growing over time without an associated cancer, and the greatest need at this time is to identify the small proportion of such cysts with PDAC and/or HGD. Until such time, judicious selection of patients for surveillance and reasonable intervals of such surveillance with selective use of EUS will help identify patients requiring resection. In our enthusiasm to detect every possible pancreatic cancer, we should not ignore the unintentional outcomes of surgery to a large majority of patients who would never develop PDAC and the astronomical costs associated with such practice.
Dr. Vege is professor of medicine at the Mayo Clinic. He reported having no conflicts of interest regarding this article.
References
Vege SS et al. Gastroenterology. 2015;148:819-22.
Lobo JM et al. Surgery. 2020;168:601-9.
Lennon AM and Vege SS. Clin Gastroenterol Hepatol. 2022;20:1663-7.
Harris RP. Ann Intern Med. 2015;162:787-9.
Dear colleagues,
Pancreas cysts have become almost ubiquitous in this era of high-resolution cross-sectional imaging. They are a common GI consult with patients and providers worried about the potential risk of malignant transformation. Despite significant research over the past few decades, predicting the natural history of these cysts, especially the side-branch intraductal papillary mucinous neoplasms (IPMNs), remains difficult. There have been a variety of expert recommendations and guidelines, but heterogeneity exists in management especially regarding timing of endoscopic ultrasound, imaging surveillance, and cessation of surveillance. Some centers will present these cysts at multidisciplinary conferences, while others will follow general or local algorithms. In this issue of Perspectives, Dr. Lauren G. Khanna, assistant professor of medicine at NYU Langone Health, New York, and Dr. Santhi Vege, professor of medicine at the Mayo Clinic, Rochester, Minn., present updated and differing approaches to managing these cysts. Which side of the debate are you on? We welcome your thoughts, questions and input– share with us on Twitter @AGA_GIHN
Gyanprakash A. Ketwaroo, MD, MSc, is associate professor of medicine, Yale University, New Haven, Conn., and chief of endoscopy at West Haven (Conn.) VA Medical Center. He is an associate editor for GI & Hepatology News.
Continuing pancreas cyst surveillance indefinitely is reasonable
BY LAUREN G. KHANNA, MD, MS
Pancreas cysts remain a clinical challenge. The true incidence of pancreas cysts is unknown, but from MRI and autopsy series, may be up to 50%. Patients presenting with a pancreas cyst often have significant anxiety about their risk of pancreas cancer. We as a medical community initially did too; but over the past few decades as we have gathered more data, we have become more comfortable observing many pancreas cysts. Yet our recommendations for how, how often, and for how long to evaluate pancreas cysts are still very much under debate; there are multiple guidelines with discordant recommendations. In this article, I will discuss my approach to patients with a pancreas cyst.
At the first evaluation, I review available imaging to see if there are characteristic features to determine the type of pancreas cyst: IPMN (including main duct, branch duct, or mixed type), serous cystic neoplasm (SCA), mucinous cystic neoplasm (MCN), solid pseudopapillary neoplasm (SPN), cystic neuroendocrine tumor (NET), or pseudocyst. I also review symptoms, including abdominal pain, weight loss, history of pancreatitis, and onset of diabetes, and check hemoglobin A1c and Ca19-9. I often recommend magnetic resonance cholangiopancreatography (MRCP) if it has not already been obtained and is feasible (that is, if a patient does not have severe claustrophobia or a medical device incompatible with MRI). If a patient is not a candidate for treatment should a pancreatic malignancy be identified, because of age, comorbidities, or preference, I recommend no further evaluation.
Where cyst type remains unclear despite MRCP, and for cysts over 2 cm, I recommend endoscopic ultrasound (EUS) for fluid sampling to assist in determining cyst type and to rule out any other high-risk features. In accordance with international guidelines, if a patient has any concerning imaging features, including main pancreatic duct dilation >5 mm, solid component or mural nodule, or thickened or enhancing duct walls, regardless of cyst size, I recommend EUS to assess for and biopsy any solid component and to sample cyst fluid to examine for dysplasia. Given the lower sensitivity of CT for high-risk features, if MRCP is not feasible, for cysts 1-2 cm, I recommend EUS for better evaluation.
If a cyst is determined to be a cystic NET; main duct or mixed-type IPMN; MCN; or SPN; or a branch duct IPMN with mural nodule, high-grade dysplasia, or adenocarcinoma, and the patient is a surgical candidate, I refer the patient for surgical evaluation. If a cyst is determined to be an SCA, the malignant potential is minimal, and patients do not require follow-up. Patients with a pseudocyst are managed according to their clinical scenario.
Many patients have a proven or suspected branch duct IPMN, an indeterminate cyst, or multiple cysts. Cyst management during surveillance is then determined by the size of the largest cyst and stability of the cyst(s). Of note, patients with an IPMN also have been shown to have an elevated risk of concurrent pancreas adenocarcinoma, which I believe is one of the strongest arguments for heightened surveillance of the entire pancreas in pancreas cyst patients. EUS in particular can identify small or subtle lesions that are not detected by cross-sectional imaging.
If a patient has no prior imaging, in accordance with international and European guidelines, I recommend the first surveillance MRCP at a 6-month interval for cysts <2 cm, which may offer the opportunity to identify rapidly progressing cysts. If a patient has previous imaging available demonstrating stability, I recommend surveillance on an annual basis for cysts <2 cm. For patients with a cyst >2 cm, as above, I recommend EUS, and if there are no concerning features on imaging or EUS, I then recommend annual surveillance.
While the patient is under surveillance, if there is more than minimal cyst growth, a change in cyst appearance, or development of any imaging high-risk feature, pancreatitis, new onset or worsening diabetes, or elevation of Ca19-9, I recommend EUS for further evaluation and consideration of surgery based on EUS findings. If an asymptomatic cyst <2 cm remains stable for 5 years, I offer patients the option to extend imaging to every 2 years, if they are comfortable. In my experience, though, many patients prefer to continue annual imaging. The American Gastroenterological Association guidelines promote stopping surveillance after 5 years of stability, however there are studies demonstrating development of malignancy in cysts that were initially stable over the first 5 years of surveillance. Therefore, I discuss with patients that it is reasonable to continue cyst surveillance indefinitely, until they would no longer be interested in pursuing treatment of any kind if a malignant lesion were to be identified.
There are two special groups of pancreas cyst patients who warrant specific attention. Patients who are at elevated risk of pancreas adenocarcinoma because of an associated genetic mutation or a family history of pancreatic cancer already may be undergoing annual pancreas cancer screening with either MRCP, EUS, or alternating MRCP and EUS. When these high-risk patients also have pancreas cysts, I utilize whichever strategy would image their pancreas most frequently and do not extend beyond 1-year intervals. Another special group is patients who have undergone partial pancreatectomy for IPMN. As discussed above, given the elevated risk of concurrent pancreas adenocarcinoma in IPMN patients, I recommend indefinite continued surveillance of the remaining pancreas parenchyma in these patients.
Given the prevalence of pancreas cysts, it certainly would be convenient if guidelines were straightforward enough for primary care physicians to manage pancreas cyst surveillance, as they do for breast cancer screening. However, the complexities of pancreas cysts necessitate the expertise of gastroenterologists and pancreas surgeons, and a multidisciplinary team approach is best where possible.
Dr. Khanna is chief, advanced endoscopy, Tisch Hospital; director, NYU Advanced Endoscopy Fellowship; assistant professor of medicine, NYU Langone Health. Email: Lauren.Khanna@nyulangone.org. There are no relevant conflicts to disclose.
References
Tanaka M et al. Pancreatology. 2017 Sep-Oct;17(5):738-75.
Sahora K et al. Eur J Surg Oncol. 2016 Feb;42(2):197-204.
Del Chiaro M et al. Gut. 2018 May;67(5):789-804
Vege SS et al. Gastroenterology. 2015 Apr;148(4):819-22
Petrone MC et al. Clin Transl Gastroenterol. 2018 Jun 13;9(6):158
Pancreas cysts: More is not necessarily better!
BY SANTHI SWAROOP VEGE, MD
Pancreas cysts (PC) are very common, incidental findings on cross-sectional imaging, performed for non–pancreas-related symptoms. The important issues in management of patients with PC in my practice are the prevalence, natural history, frequency of occurrence of high-grade dysplasia (HGD) and/or pancreatic cancer (PDAC), concerning clinical symptoms and imaging findings, indications for EUS and fine-needle aspiration cytology, ideal method and frequency of surveillance, indications for surgery (up front and during follow-up), follow-up after surgery, stopping surveillance, costs, and unintentional harms of management. Good population-based evidence regarding many of the issues described above does not exist, and all information is from selected clinic, radiology, EUS, and surgical cohorts (very important when trying to assess the publications). Cohort studies should start with all PC undergoing surveillance and assess various outcomes, rather than looking backward from EUS or surgical cohorts.
The 2015 American Gastroenterological Association guidelines on asymptomatic neoplastic pancreas cysts, which I coauthored, recommend, consistent with principles of High Value Care (minimal unintentional harms and cost effectiveness), that two of three high-risk features (mural nodule, cyst size greater than 3 cm, and dilated pancreatic duct) be present for EUS-guided fine-needle aspiration (EUS-FNA). By the same token, they advise surgery for those with two of three high-risk features and or concerning features on EUS and cytology. Finally, they suggest stopping surveillance at 5 years if there are no significant changes. Rigorous GRADE methodology along with systematic review of all relevant questions (rather than cohorts of 500 or fewer patients) formed the basis of the guidelines. Those meta-analyses showed that risk of PDAC in mural nodules, cyst size >3 cm, and dilated pancreatic duct, while elevated, still is very low in absolute terms. Less than 20% of resections for highly selected, high-risk cysts showed PDAC. The guidelines were met with a lot of resistance from several societies and physician groups. The recommendations for stopping surveillance after 5 years and no surveillance for absent or low-grade dysplasia after surgery are hotly contested, and these areas need larger, long-term studies.
The whole area of cyst fluid molecular markers that would suggest mucinous type (KRAS and GNAS mutations) and, more importantly, the presence or imminent development of PDAC (next-generation sequencing or NGS) is an exciting field. One sincerely hopes that there will be a breakthrough in this area to achieve the holy grail. Cost effectiveness studies demonstrate the futility of existing guidelines and favor a less intensive approach. Guidelines are only a general framework, and management of individual patients in the clinic is entirely at the discretion of the treating physician. One should make every attempt to detect advanced lesions in PC, but such effort should not subject a large majority of patients to unintentional harms by overtreatment and add further to the burgeoning health care costs in the country.
PC are extremely common (10% of all abdominal imaging), increase with age, are seen in as many as 40%-50% of MRI examinations for nonpancreatic indications, and most (>50%) are IPMNs. Most of the debate centers around the concerns of PDAC and/or HGD associated with mucinous cysts (MCN, IPMN, side-branch, main duct, or mixed).
The various guidelines by multiple societies differ in some aspects, such as in selection of patients based on clinical, laboratory, and imaging findings for up-front surgery or surveillance, the frequency of surveillance based on the size of the cyst and the presence of other concerning cyst features (usually with MRCP), the indications for EUS (both initial and subsequent), importance of the magnitude of growth (most IPMNs slowly grow over a period of time), indications for surgery during surveillance and postsurgery surveillance, and the decision to stop surveillance at some point in time. The literature is replete with small case series reporting a proportion of cancers detected and often ignoring the harms of surgery. Incidence of and mortality caused by PDAC are very low (about 1% for both) in a large national cohort of VA pancreatic cyst patients with long-term follow-up and other studies.
Marcov modeling suggests that none of the guidelines would lead to cost-effective care with low mortality because of overtreatment of low-risk lesions, and a specificity of 67% or more for PDAC/HGB is required. AGA guidelines came close to it but with low sensitivity. Monte Carlo modeling suggests that less intensive strategies, compared with more intensive, result in a similar number of deaths at a much lower cost. While molecular markers in PC fluid are reported to increase the specificity of PDAC/HGD to greater than 70%, it should be observed that such validation was done in a small percentage of patients who had both those markers and resection.
The costs of expensive procedures like EUS, MRI, and surgery, the 3% complication rate with EUS-FNA (primarily acute pancreatitis), and the 1% mortality and approximately 20%-30% morbidity with surgery (bleeding, infection, fistula) and postpancreatectomy diabetes of approximately 30% in the long run need special attention.
In conclusion, one could say pancreas cysts are extremely frequent, most of the neoplastic cysts are mucinous (IPMN and MCN) and slowly growing over time without an associated cancer, and the greatest need at this time is to identify the small proportion of such cysts with PDAC and/or HGD. Until such time, judicious selection of patients for surveillance and reasonable intervals of such surveillance with selective use of EUS will help identify patients requiring resection. In our enthusiasm to detect every possible pancreatic cancer, we should not ignore the unintentional outcomes of surgery to a large majority of patients who would never develop PDAC and the astronomical costs associated with such practice.
Dr. Vege is professor of medicine at the Mayo Clinic. He reported having no conflicts of interest regarding this article.
References
Vege SS et al. Gastroenterology. 2015;148:819-22.
Lobo JM et al. Surgery. 2020;168:601-9.
Lennon AM and Vege SS. Clin Gastroenterol Hepatol. 2022;20:1663-7.
Harris RP. Ann Intern Med. 2015;162:787-9.