Thank you, Michelle. Before I introduce our next speaker, I want to tell you a bit about Dr. Basil Hershowitz. Dr. Hershowitz is truly a principal figure in the field of GI endoscopy, and he's credited with inventing the first fully flexible fibroptic endoscope. ASGE honors the late Dr. Hershowitz for his leadership and accomplishments through this annual lectureship. This lecture is presented by the ASGE Foundation and was established through the generous support of ASGE as well as friends, trainees, and colleagues of Dr. Hershowitz. Therefore, it's my great honor to introduce our speaker for the 2021 Basil Hershowitz lecture, Dr. Grace Elta. Dr. Elta is professor emeritus at the University of Michigan, coincidentally where Dr. Hershowitz worked when he developed the fibroptic endoscope. Grace has served many roles in the ASGE over the years, including as secretary and president, and in 2018, she was awarded our society's highest honor, the Schindler Award. Also in 2018, she was the lead author on the ACG guideline on the diagnosis and management of pancreatic cysts. So please join me in welcoming Dr. Grace Elta to present this year's Basil Hershowitz Endowed Lecture, where she'll be discussing pancreatic cysts. I want to start out by thanking Klaus Mergener, Jason Donowitz, and also Rob Enns for the honor of giving this lecture, the Basil Hershowitz Lecture. I want to show you this picture here. This actually adorns the hospital-based unit at Michigan Medicine because Dr. Hershowitz was actually at Michigan when he did his pioneering work in fibroptics in conjunction with the Department of Engineering. My disclosure is Olympus Consulting. So diagnosis of pancreatic cysts. Most pancreatic cysts are detected on imaging for non-pancreatic indications. The prevalence in asymptomatic people varies widely in the literature, although includes all ages, and therefore is anywhere from 2.4 to 13 percent. However, when MRI is done for non-pancreatic indication in patients older than 70, 40 percent have incidental pancreatic cysts. Luckily, cysts larger than two centimeters has a much lower incidence, and those are the ones, of course, that are more concerned. This increased recognition, which we've seen over the last decade, is likely due to both use of more and better quality imaging, not due to increased incidence. Oops. So pancreatic cysts, why all the hype? Some pancreatic cysts are precursors of cancer, although the vast majority do not progress. So the key question is, what is the cancer risk? I've quoted two more recent studies below. One looked at low-risk IPMNs versus high-risk IPMNs and found that at 10-year follow-up, low-risk IPMNs have a cancer risk of 7.7 percent, whereas in high risk it's 24.6 percent. A second study of over 1,400 patients followed for 20 years noted that pancreatic cancer was seen in 3.3 percent at five years, 6.6 percent at 10 years, and 15 percent at 15 years. Even small cysts, and they define small as less than 15 millimeters in the study, had sevenfold increased risk of pancreatic cancer compared to the general population. 44 percent of those cancers were concomitant ductal adenocarcinomas, 56 percent were IPMN-derived carcinomas. This highlights the point that IPMNs are a marker for a high risk of pancreatic cancers that are not IPMN-derived also. Issues impacting pancreatic cyst management decisions. Frequent detection with a low risk of cancer. There's a high cost of cyst surveillance. Benefits of surveillance are still unproven. We really don't know if cancers can be prevented or detected early. Risk of pancreatic surgery is quite high. Mortality rate, even in expert centers, is about 2.1 percent, and morbidity is as high as 30 percent. Also, cysts are often found in elderly patients who have many comorbidities. Now I want to talk briefly about types of cysts. IPMN, or introductal papillary mucinous neoplasm, is by far the most common cyst that are diagnosed incidentally. Usually these are side branch. IPMNs have an equal incidence in men and women, often diagnosed in the seventh decade. If you do an aspirate, the fluid is high in CEA and high in amylase and lipase. The side branch ones communicate with the main pancreatic duct, and in this image up here you can see a side branch one right here. 40 percent of these are multifocal, as is seen in this particular image. And as I mentioned earlier, this is by far the most common incidental cyst we see. The main duct is much less common than side branch and has a high risk of cancer. In 50 percent of patients, there's a patchless papilla with mucin protruding, as you see in this image. The mixed IPMNs, which is a combination of side branch and main duct, have the same cancer risk as main duct. Another common pancreatic cyst is the serous cyst adenoma. 75 percent of these occur in women, usually in diagnoses made in the sixth decade. There is a 0.1 percent risk of serous cyst adenoma, adenocarcinoma, I'm sorry, but because this is so low, these cysts do not require surveillance. They rarely cause symptoms, although when they're very large and they compress adjacent organs like the bile duct, they can cause symptoms. The majority of these are microcystic or honeycomb, as seen in the CT scan. The microcystic ones can look solid on CT, although MRI or EOS, they look more cystic. The oligosystic ones are less common, but are more difficult to differentiate between side branch IPMN. Central scar is present in less than 30 percent, but is seen in this image here with the calcified central scar there, and that's fairly diagnostic of a serous cyst adenoma on imaging. If you do aspirate, they're low in CEA and low in amylase or lipase. Pseudocysts are the only non-neoplastic cysts that I'll mention. It occurs in acute and chronic pancreatitis. If you do aspirate, the fluid is often brown with a high lipase and a low CEA. When it's mostly debris, we call it walled-off pancreatic necrosis. As you can see in this image, there's acute pancreatitis with a lot of inflammation around there and a pseudocyst here. The most important point on this slide is the caution at the bottom, and that is that neoplastic cysts can cause pancreatitis, so a cyst at initial presentation requires investigation. I'll briefly mention some of the less common pancreatic cysts. One of those is the mucinous cystic neoplasm, or MCN. MCNs are the other mucinous-type cysts besides the IPMMs. They're almost all seen in women, usually in the fifth to seventh decade. They're more commonly seen in the body or tail, and on histology have ovarian stroma on pathology findings. They do produce mucin, but they do not communicate to the main pancreatic duct like the common side branch IPMM. If you do an aspirate, it's high in CEA and variable in lipase. The risk of cancer is lower than we previously thought. I give you two case series here. They're rare enough that we only have case series data. In 90 resected MCNs, only 10% had cancer or high-grade dysplasia. In the second case series, which was 344 MCNs, they were all less than three centimeters in size, and none of them had a solid component, but there was no high-grade dysplasia or cancer. Another less common pancreatic cyst is the solid pseudopapillary neoplasm. The ratio is very high, women to men 10 to 1. It can occur in any part of the pancreas, and occurs most commonly in the third decade, so in young 20-year-old women, although there is a wide age range for this particular cyst. Small ones can be more cystic without much cystic, more solid without much cystic degeneration. This lesion requires surgical removal. There's aggressive histologic tumor behavior in about 10%, although the five-year disease-free survival in a case series published in 2014 was 98%. Another less common pancreatic cyst is the cystic neuroendocrine tumor. It's usually non-functioning, and these patients may have MEN1 syndrome. It's equally common in men and women, most commonly seen in the fifth to sixth decade. If you do an aspirate, it's low in CEA and high in lipase. EOSFNA is often needed for diagnosis, and luckily the cytology is usually positive. There are other rare pancreatic cysts, simple cysts, simple epithelial cysts, cystic degeneration of ductal cancer, and hidatid cysts, and I won't talk further about those. Important question on this slide is, how do we determine what type of cyst it is? MRCP is thought to be the best choice because it's non-invasive, it has no radiation, and it's somewhat better than CT scan for pancreatic duct connection. CT or EOS without FNA is an excellent alternative, although you should remember that EOS is operator dependent. Indeterminate cysts by one imaging modality may benefit from a second modality, and that has been shown in the literature. Accuracy of all three modalities is fairly similar, MRI, CT, EOS imaging without FNA. Accuracy for cyst type in K-series is 40 to 50 percent. Accuracy for determining benign versus malignant in K-series is 55 to 76 percent. When should you add FNA or micro forceps biopsy to assess the cyst type? The answer is when the results will alter your management for indeterminate cysts. Recall that when you add FNA or biopsy, you increase risk of the procedure, and therefore you don't want to do that unless you're going to change your management. The majority of pancreatic cysts found on imaging do not require an FNA. The caveat point on this slide is that the diagnostic accuracy of MRCP, CT, and EOS is relatively low and somewhat disappointing. I don't know. There it goes. For some reason it didn't want to advance. Other cyst sampling techniques. There are other molecular markers that can be measured. KRAS and GNAS are the common ones. A meta-analysis of multiple studies show that the sensitivity and specificity for diagnosing a mucinous cyst is better for KRAS and GNAS than simple CEA. However, those tests are somewhat costly, and they do not help determine cancer risk. They only tell you whether it's a mucinous cyst or not. Many new markers are under study, and I've given you some references below. Next-generation sequencing, methylated DNA markers, and even simple glucose. A meta-analysis of microbiopsy forceps, which is performed through a 19-gauge FNA needle versus traditional FNA, have shown that there's improved sample adequacy with the forceps with an odds ratio of 4.8. There's also improved sample accuracy with an odds ratio of 3.4, with accuracy range anywhere from 75 to 83 percent. Unfortunately, biopsy has more complications. The complication rate in various series is running about 6 percent. The standard FNA, it's only 2 to 3 percent. Confocal microscopy is another alternative. This is a probe that's also passed through a 19-gauge FNA needle. Unfortunately, microscopy is much harder for GIs to learn, because you really have to learn cytology. But it is better than CEA and cytology in diagnosing mucinous cysts, and carries with it, at least in one case series, a 3.5 percent pancreatitis rate. Cyst surveillance recommendations. First of all, who should be surveyed? Surgically fit patients who have presumed IPMNs or MCMs. You notice I say presume, because sometimes, as we mentioned before, we're not 100 percent certain of the cyst type on imaging. How should we perform the surveillance? Well, MRCP is thought to be the preferred modality of surveillance, because of the lack of radiation and the lack of need for sedation. EOS is a great choice in patients who cannot or choose not to have MRIs. Surveillance intervals are primarily based on size. We know from 5-surgical series that cysts greater than 3 centimeters have anywhere from 4 to 47 percent chance of high-grade dysplasia or cancer. We also know, unfortunately, that some cysts less than 3 centimeters do have cancer. Therefore, size is not a great criteria, but it's the one that's commonly used. The study with the graph to the right there does show that initial cyst size correlates with development of worrisome features. On this axis is worrisome features, a chance of developing them. The worrisome features are listed up here, size greater than 30, neural nodule, cyst wall thickening, and main pancreatic duct dilation. Over here is initial cyst size, less than 10 millimeters, 10 to 20 millimeters, 20 to 30 millimeters, or over 30 millimeters. You can see that over follow-up, you're much more likely to develop worrisome cyst features if you have a larger size to begin with. Therefore, size is the criteria that's most often used. What sizes do we recommend? Well, the ACG guideline published a couple of years ago recommended MRI and MRCP every two years for cysts that are less than a centimeter, for cysts that are 1 to 2 centimeters, MRI every year, and for those that are 2 to 3 centimeters, MRI or EOS every 6 to 12 months. Some experts like to alternate MRI and EOS, thinking they get different images from those. When should you get worried when you're following a cyst? Consider shorter interval imaging or even performing an FNA or biopsy if the patient has developed new onset or worsening diabetes. That's a worrisome finding. The other finding that we look for is increasing cyst size. The ACG guideline recommends that increasing cyst size greater than 3 millimeters per year is a worrisome finding, whereas the European guidelines say greater than 5 millimeters per year. Patients who have any of the following high-risk factors should be referred to a multidisciplinary clinic and or you can perform FNA or biopsy. A cyst that causes jaundice or acute pancreatitis, an elevated serum CA99, a mural nodule or any solid component to the cyst, main pancreatic duct dilation greater than 5 millimeters, focal dilation or a change in caliber of the duct with upstream atrophy of the gland, cyst size greater than 3 centimeters for the ACG guidelines, and the European guideline recommends cyst size greater than 4 centimeters. So when should you stop surveillance for the mucilus pancreatic cyst? Recall that we're surveying IPMNs, side branch IPMNs, and MCMs. The risk of malignant transformation does not decrease with time. Multiple papers have now shown that, so there's no justification for stopping surveillance after five years. Stop if the patient is not a surgical candidate due to their age, their wishes, or their comorbidities. The Charleston Comorbidity Index is helpful in shared surveillance decision making with patients. In general, we recommend stopping between the age 76 and 85. Which pancreatic cysts require surgery? Main duct IPMNs or mixed IPMNs, and patients are good surgical candidates should be referred for surgery. Side branch IPMNs and MCMs that have high risk features in good surgical candidates, solid pseudopecular neoplasms, cystic neuroendocrine tumors that are greater than 2 centimeters, and cirrhosis adenomas that are causing symptoms due to compression of adjacent organs. So in conclusions, the majority of incidental pancreatic cysts are side branch IPMN. IPMNs and MCMs have potential for malignancy and require surveillance. MRCP is usually recommended for cyst diagnosis and surveillance, although both CT and EOS are reasonable alternatives. Cysts may require EOS, find elastoporate or biopsy for unclear diagnosis or worrisome characteristics. The choice between FNA and biopsy really depends on whether you're willing to take an increased risk for a higher yield with the biopsy forceps. Pancreatic cysts are very common and the benefit of surveillance is still not proven. Pancreatic surgery unfortunately still has significant morbidity and some mortality. Patients not fit for surgery should not have any further evaluation. Thank you very much.

pancreatic cysts

diagnosis

cancer risk

management

surveillance