Hi, this is Elliot Fishman, and welcome to our latest podcast. And this is going to be a look at a topic that's of great interest to all of us, something we talk about on a daily basis, something that's undergoing lots of change but very important to our practices, and that's going to be looking at cystic pancreatic masses.

Now, we know that cystic lesions are very common. We wrote this article a long, long time ago, at 16-slice CT, where we noted cysts seemed to be becoming more frequent, and even at 16-slice CT, just under 3% of patients had incidental cysts. And at the same time we were doing CT, MR even mentioned maybe 20% of patients had cysts. There are other articles written. As scanners get better, remember, the smaller cysts are best seen on higher-resolution scanners, whether they're CT or MR, and people talk about how over age 50, if you look really hard enough, 20% to 30% of patients will have cysts on CT, and more than 30% on MR, so they're very, very common. The question is, what do you do about it, when do you do something about it, and when do you not worry about it?

Well, when I first started CT, and all of us did, the most common cystic lesion we saw was a pseudocyst. Remember, one of the earliest applications of CT was the pancreas, particularly pancreatitis. And usually, you had a clinical history of pancreatitis, the gland would be swollen. Pseudocysts can be at time of presentation, or could be a sequela of the pancreatitis. Pseudocysts could resolve over time, but sometimes they didn't. And whenever we saw an incidental cyst, I remember way back when we would say, "Oh, patient must have had prior pancreatitis."

Some features about pseudocysts people talk about: smooth external contour, protrusion into peripancreatic soft tissues, peripancreatic stranding, and of course, a prior history of pancreatitis.

We talk about peripancreatic fluid collections, well-defined, usually several weeks post-symptoms beginning, again you can see changes in the gland, up to 20% of cases, half resolve spontaneously, other patients get pain and infection, with common locations from the lesser sac to the anterior pararenal space, left and right. And we know pseudocysts can track anywhere; they can track down into the pelvis. We saw cases with posterior mediastinal masses which were pseudocysts, or scrotal masses.

And so, a typical patient like this with pancreatitis, which had resolved, has a cystic lesion: pseudocyst. Again, most likely it's going to resolve; if it doesn't resolve, most patients will still be asymptomatic, and you will not do anything about it.

Again, the appearance: water density, well-defined, sharp margins, the vessels occasionally can be compressed if pseudocysts get large enough. Most of the time, the vessels look okay.

Pseudocysts occasionally, though it's rare like this case, can calcify over time. In patients with repeated episodes of pancreatitis, you may see pseudocysts and glandular calcifications. Chronic pancreatitis with pseudocysts. This appearance of rim calcification, I always comment when I see rim calcification: liver, spleen, kidneys, pancreas, I always worry about old trauma, but pseudocysts, like pancreatitis, can also have that rim calcification, which you can see very nicely on the MIP imaging, kind of circumferential around the lesion.

Now, we typically don't worry about pseudocysts, right? They're going to resolve. At times they don't resolve, at times they'll need to be drained, but typically, that's not the problem we have, and that's not really where we're seeing all the incidental pancreatic cysts as we've learned over time. Really, what we're talking about is IPMNs: intraductal papillary mucinous neoplasms. It's usually in older population, seventh decade, a bit more common in men.

The key to look at is whether it's a cystic lesion, what is its size (under 2 cm, under 3 cm), and then where is it going. All pseudocysts—or all IPMNs, rather—will attach to the pancreatic duct. It's sometimes hard to see that, it's a bit easier to see on MR than CT, and we classify lesions as side-branch (that's the majority). We talk about those with a main duct dilated, which usually means a duct of 7 mm or greater, and a mixed type where there's a dilated duct as well as a side-branch component. You can see one IPMN or you can see multiple IPMNs. When patients have a main duct that's dilated, again, now 7 mm (not the 1 cm we used to speak about), those are more concerning because main-duct IPMNs have a higher incidence of malignancy and usually require surgery.

When you look at the whole IPMN population, if you look over someone's entire life, there's less than a 3% chance those patients could develop tumors of malignant course. Now, when you think about IPMNs, what are predictors of malignancy? Well, lesion size, over 3 cm. Remember, the Tanaka criteria would be that if it's over 3 cm cystic lesion, you resect it. We talk about interval growth of greater than 2 mm a year. We talk about mural nodule or nodules, particularly enhancing nodules. We talk about concern when there are thick septations, again, enhancing septations are more concerning, and we always worry when patients have clinical symptoms, including abdominal pain and unexplained episodes of pancreatitis. When patients have cystic lesions that are a bit concerning, and they have symptoms that you can't explain by pancreatitis or other reasons, those patients invariably are going to have resection.

We also talk about patients with new-onset of diabetes. There's a lot of work going on, and I don't have time to discuss it today, but patients with new-onset diabetes, you worry about: is diabetes a sign that the patient has early pancreatic problems, and that lesion is going to be of concern?

So when you look at this, this was a good article from a while back. You see a normal duct, and then we talk about PanINs, and PanINs is one of the precursors, though infrequent, of pancreatic ductal adenocarcinoma. And we also talk about IPMNs and MCNs as macroscopic precursors. So again, it's this idea about maybe you don't have a tumor today, but what's going to happen in the future? And that really becomes the challenge for us because, what do we do? If you operate on everybody with a cystic pancreatic lesion, you're talking about 30% or more of the population over age 50. That becomes very, very, okay, it's unreal, you can't do that. Obviously, patients with pancreatic surgery, there's all sorts of complications. And also, if you have an IPMN and you take it out and it's low-grade dysplasia, it doesn't mean you're not going to get another IPMN. If you really wanted to take care of patients and remove the chance of having a tumor eventually, you would have to do a total pancreatectomy, and total pancreatectomies have tremendous complication rates over time, patients are often brittle diabetics, and there's all sorts of issues. Not to forget also that pancreatic surgery has a mortality rate in general hands of between 2% and 5%. So again, it's not a simple walk in the park.

Now, in this article by Wang, up to 15% of pancreatic cancer are thought to arise from precursor mucinous lesions, namely IPMNs and MCNs. IPMNs are further classified, as we mentioned, branch, main duct, mixed-type IPMNs. Due to higher means of rate of malignancy for resected main-duct lesions, they're the ones that are going to be removed with greatest concern. Branch-duct IPMNs typically undergo surveillance because they are prevalent in the general population, yet carry a much lower rate of malignancy. And so, a lot of the work, and I'll mention it over the next few minutes, is how do you manage a patient with an incidental IPMN, a side-branch lesion? Who do you follow? Who do you do ERCP? Who do you do EUS? Who do you resect? Do we have an answer?

I mentioned about the guidelines, two main criteria to predict malignancy: the worrisome features (cyst over 3 cm, solid nodule, cyst wall thickening, change in main pancreatic duct caliber, glandular atrophy, and a fast growth rate), and all of those make sense because that seems like the lesion is rapidly changing. When a lesion is rapidly changing, it's growing, changing shape, you really have to worry. These findings lead to a EUS, and then you will get tissue sampling. Imaging findings of jaundice, enhancing mural nodule, main duct over 10 mm, that's kind of easy because high risk, high likelihood of malignancy.

Some examples: Here's just a simple case of an incidental under 2 cm IPMN. Now, you don't see the communication on these images with the pancreatic duct; it may be hard on CT unless you really play around with it, it's a little bit easier on the MIP MRs, but that's a lesion, there's no septations, it's small. Depending on the patient's history and family history and risk factors, you might do nothing and say, "I'm not worrying about it," or maybe bring the patient back in a year. If the patient had high risk, family history, or patient had BRCA2, you might consider doing an EUS, but that's kind of a little bit of overkill, probably you'll go to MR, it's non-invasive, and if it looks okay, stop at that point, and then maybe schedule a follow-up for a year or so later.

Here's another one: you can see the pancreatic duct, and you can see a 1 cm cystic lesion by the tail of the pancreas.

Now, I mentioned you can have multiple lesions. Here's one in the uncinate, here's one in the body, another one in the body, one in the tail. And that becomes more challenging because you have to evaluate each of the lesions. You may have five lesions, four look perfect, one looks maybe abnormal. But what are you going to do? Again, remember the comment I made: if you really wanted to, quote-unquote, prevent cancer and cure the patient, you would have to do a total pancreatectomy. Just sampling one lesion, you can't assume that the other four lesions are going to be exactly the same.

Now, sometimes it's tricky. I'm not sure if you have multiple lesions or septations. In this case, there's a cystic lesion head to uncinate, but there are several small septations present, or small lesions. These in fact ended up being two lesions kind of kissing each other rather than a central septation. Here's another set of images where you can see on the coronals, these are septations. Sometimes the diagnosis of septations is easier to see on the coronal images. Thin septation does not increase the risk of tumor. Now, we talk about septations when we'll speak about MCN, once they have thickened septations, MCNs as well, then you worry about the lesion being at least a moderate or high-grade dysplasia.

So here's another lesion. This one was over 3 cm, patient had some symptoms so it was resected. This was an IPMN with an intermediate-grade dysplasia. So, patient was basically cured. You would still follow this patient to make sure no other lesions develop in the pancreas.

Here's another patient with a larger lesion, thicker septations. This was an IPMN with dysplasia, in fact high-grade dysplasia. This could be an MCN. Location of MCNs, and we'll discuss that, body of the pancreas. So, I would have read this as concerning lesion, I would have favored an MCN, but it was an IPMN. And on the coronal view, you really get the feel of the size of the lesion, the septations, and you can see why everyone would have been concerned and this lesion would have been resected.

Now, when you have a lesion like this where the main pancreatic duct is dilated to 17 mm, remember we said 7 or better we worry, 17 like this, there's soft tissue mass infiltrating the duct. This was only moderate dysplasia. This patient needs a total pancreatectomy. I would have guessed the patient had high-grade dysplasia, but you can see markedly dilated duct, thickening of the wall, septations, all sorts of prominence, you know, but this was an intermediate grade. Here is the same case on the 3D imaging, volume rendering, again thickened wall irregularity. This patient's going to get a pancreatectomy because, in my mind, that looks like a tumor, and it's a main-duct IPMN, and all main-duct IPMNs are going to get resected, and with the thickened wall and everything else, it's particularly worrisome.

Now, the idea of how to manage IPMNs is really a challenge. And trying to figure out a way, the ACR had these guidelines. Now, the challenge is if you come to certain institutions, a Hopkins, a Mass General, an NYU, Mallinckrodt—I'm just going across the country—M.D. Anderson, UCSF, Stanford, one of the challenges is that not everyone follows the same guidelines because there are so many guidelines. Everybody has local expertise, and they're doing the best they can. So what was tried to be done was come up with something that cuts across everybody.

Now, Alec Megibow is very, very smart guy in pancreas. So what he did, he tried to break things up into categories: under 1.5, 1.5 to 2.5 cm incidental pancreatic cyst with MPD communication, 1.5 to 2.5 cm incidental pancreatic cyst without MPD communication or cannot be determined, over 2.5 cm incidental pancreatic cyst, and an incidental pancreatic cyst in patient over 80 years old. They recommended 9- to 10-year follow-up till age 80. But again, you can see that 9- to 10-year follow-up, that's like a lifetime, and why do you stop at 10? If you stop at 10, there's still the same risk, so why are you stopping at 10? It's almost like you're saying, "Well, if a patient is over a certain age, they're not going to be a candidate for surgery." That makes sense, but to say if you followed for 10 years and you don't find anything, it's negative, that's not true. The risk is still there. It's not like looking at lung nodules where you stop at a year or two years, right?

So, we have these complicated different charts. So let's do this: we've covered a lot of material, let's come back to Alec's look at these lesions, what he was thinking, what we can learn, and what we could do.

So let's take a break and come back in a few minutes.