Dr. Rubin discusses Dysplasia in Inflammatory Bowel Disease: It’s More Than Meets the Eye.
[MUSIC PLAYING] DAVID T. RUBIN: We're going to talk about dysplasia and IBD. And there's actually been quite a bit of progress in this area of our field. I want to bring you up to speed, because what we have been doing for years is now changing, and needs to be considered. So I'll talk to you a little bit about the newer data regarding risk factors and prevention strategies. We'll talk about a newer concept of stratifying your UC patients who have dysplasia, to determine who might benefit from surgery versus those who may be safely followed. And then I'll talk to you a little bit more about where the field is trying to go. So before we talk about any cancer prevention, we have to understand what are the factors that lead to the cancer itself, and how might we intervene to actually change outcomes. That includes understanding the risk factors, knowing which patients are at highest risk. Having the tools to detect pre-cancer, so that you can impose a change upon the patient, or on the disease process, so that you can modify the subsequent outcomes of interest. In ulcerative colitis, as much as we say we'd like to, of course, save lives, we're focused more often in the day to day discussions about all this, about trying to save a colon. And, of course, we recognize that often that's not the right decision, and we should be moving our patients towards surgery. Understanding, as well-- and a topic that's not in this lecture-- is the outcome of the surgery and quality of life issues. But let's focus on cancer risk, and how we work to prevent those. Now in this now historic meta-analysis, it looked at all the old studies of colorectal cancer and ulcerative colitis. And these studies were predominately in the era before we had a knowledge of all the risk factors. And these studies were also in the era before we were really using standard surveillance and prevention strategies. So what we were quoting when I was finishing my fellowship in 2001, was that people who have chronic ulcerative colitis had a 2% risk of colon cancer by the time they'd had it for 10 years. That risk jumped to 8% by 20 years, and was as high as almost 20% by 30 years of disease. As many in the room may recall, it used to be the recommendation that after you had colitis long enough, you should schedule an elective colectomy, because we didn't know how to prevent cancer and the risk was just considered to be too great. Now, subsequently, there have been a lot of studies and some more recent analyses, including this one which is from a population-based cohort in the Scandinavian area that suggests that, in fact, the risk of cancer in colitis and in Crohn's disease of the colon was not greater than the general population. Now, I want to point out that that particular population has previously demonstrated some of these results which are different than what we've seen in North America and other parts of Europe. Now, one way to explain the discrepancy between the historical meta-analysis and that population-based study would be if we did a more recent meta-analysis like this one that was published, which separates out those patients who've been seen and have been analyzed in referral centers like ours, and those patients who might be assessed in more of a population-based analysis; in which such risk factors like severity of inflammation, family history, exposure to therapies, are not controlled. And you can see here that referral centers-- where there's more accurate data, but admittedly a biased population-- certainly have a much higher rate of fighting dysplasia and colorectal cancer than those who were in population-based studies. So in general in our field, we still acknowledge-- and this is sort of intuitive-- that chronic inflammation does increase the risk of cancer, and a preventive strategy is warranted. So our updated risk factors for dysplasia and colorectal cancer and ulcerative colitis-- and we extrapolate with less data to Crohn's colitis-- include, of course, the duration of disease, a greater extent of chronic involvement. Increased inflammatory activity is now confirmed as an independent risk factor, and a very important one. A family history of colon cancer, separate from a family history of IBD. PSC remains a potent, independent risk factor. A younger age of diagnosis, independent of disease duration. Backwash ileitis has been associated, as have pseudo-polyps, and both may actually just be surrogates of more severe or extensive inflammation, rather than true independent risk factors. And, of course, a mass or stricture in the colon has been associated with neoplasia. Prior dysplasia is an important risk factor for future dysplasia. And, very importantly-- and sometimes ignored-- is the fact that males have a higher risk than females. And we should keep that in mind when we stratify risk-prevention strategies in our follow up. Our prior, and still current in the US, guidelines for cancer prevention, you see, as well as Crohn's colitis, are actually out of date. So I am loath to tell you about them, only to show you where we're having problems in using them currently. We say that after eight to 10 years, somebody should enter a prevention strategy or program. The interval for follow-up varies, and there's no real definition of how to do that. So anywhere between one and three years in follow-up. From old, retrospective analyses with poor technology, we say you need to do at least 33 biopsies randomly throughout the colon to adequately and systematically sample the mucosa in order to be comfortable that, if there's no dysplasia, you haven't missed any. There's a lot of problems with that. But that's the best we used to be able to do. Chromo, which we acknowledge is superior-- and I'll show you a little data-- is not yet recommended as standard of care, but it is acknowledged to be better than random biopsies. And the question remains, is it something we should be doing more often, and in whom? And I'll show you some data on that. For those patients with confirmed high-grade dysplasia, there remains little debate about performing a colectomy. Although there's one study in our own data which suggests that high-grade dysplasia that's confined to a polypoid lesion with no other problems in that patient might be followed with more intensive surveillance. But those patients who have low-grade dysplasia, either polypoid lesions, or even now low-grade dysplasia that's found in flatter areas, may be followed with vigilent follow-up, which I'll define for you in a moment. And, of course, as always the unresectable lesions with dysplasia require surgery. Now, there have been some challenges to these guidelines, and they're based on the fact that our technology has evolved, and we've learn more about the natural history of cancer, and neoplasia, and ulcerative colitis and Crohn's. First one is that we acknowledge that this random biopsy approach that we've been doing all these years, first of all, wasn't even being followed by many gastroenterologists around the world. And, secondly, was a very low yield. And in two different retrospective analyses, including one from our center, we realized that even with the older scopes using white light, most neoplasia was actually visible. And, in fact, if you did a per patient analysis, you would find that every time there was a cancer, there was something that was seen on a white light colonoscopy. So it's not that we've been missing all sorts of lesions all this time, such that we need to embrace all the new technology. Even with the older technology, and with retrospect, we can see a lot of things that we were looking for with dysplasia. So when you combine these two observations and some additional studies, you realize that, in fact, we have been able to see much more than we were acknowledging, and that random biopsies may be a waste of time and money. So what's the approach to visible dysplasia that we recommend? This is a review that I wrote with Jerry Turner, who's one of our GI pathologists. And it first acknowledges that the old term which was coined at the University of Chicago, of DALM-- which people still throw around-- the dysplasia-associated lesion or mass. Or ALM-- which was, I think, developed by Charles Bernstein in Canada-- the adenoma-like mass or lesion, are being replaced by more specific descriptive terms, like a polypoid lesion, or a non-polypoid lesion, or a flat lesion, or invisible dysplasia-- meaning it was found on a random biopsy, and you didn't see anything. If the lesion is visible and raised, and a complete endoscopic resection can be performed, more vigilant and close follow-up is recommended and can be performed appropriately. Now the caveats to all this, of course, are is the patient under good control from their inflammation, will they follow up the way you're recommending, and what are their other risk factors. However, if the lesion is flat, and it's low-grade and not multifocal, you might still consider follow-up with the patient-- again, with all those caveats-- but multifocal lesions or high-grade dysplasia should prompt a colectomy, and send it to your surgical colleagues. So how are we updating our surveillance approach with the newest technologies? High-definition scop, chromoendoscopy, and some other approaches. Well, first, I'll acknowledge that there have been multiple tandem colonoscopy studies that compare white light to chromoendoscopy with dye spray-- whether it's methylene blue or indigo carmine. And consistently across these studies has been the finding that the chromo approach identifies more dysplasia than white light. So on the one hand, you might say, well, that's a no brainer. We should be doing this in everybody. However, it really depends on what you're looking for, and whether or not it changes your outcomes, if we should actually invest in the time and skills necessary to do this properly. Because you might say, well, if we were missing all this dysplasia all these years by not doing chromo, where are all the occult malignancies that should have been appearing before our eyes? So it might be that we're actually finding dysplasia that doesn't have the same natural history of the dysplasia of yesteryear, or the dysplasia that was associated with all the bad outcomes of the old studies. So that's one possible explanation. In addition, it's possible that in the current era of medical management-- that Atsushi reviewed so nicely-- we're actually changing the natural history of dysplasia, and there's less progression than we realized. So the only study that's had any follow-up from chromo was the one done in Mount Sinai in New York, by Jim Marian and colleagues. And of the patients who actually went to surgery, there were no cancers found, and admittedly a small number, and we certainly need more information. But all these other studies that tell us that the technique identifies more dysplasia, have not yet actually demonstrated that it changes our outcomes in any appreciable manner. It's going to be very hard to do a large study to actually figure that out. Now the formula that I've been using for a while is with methylene blue. I prefer methylene blue, it's an absorptive dye. Indigo carmine is a surface dye. And I liked the absorptive capabilities. But our interventional endoscopists usually like indigo carmine, because they don't like things that are absorbed when they're trying to just identify lesions for removal. But when I'm spraying the whole colon, I don't want to have to rely on gravity and moving the patient around for a surface dye. There are some data in animals that suggest methylene blue may be carcinogenic, but the concentration we're using in humans for these intermittent exams is very low, and that's not been a concern of ours overall. I take two vials of 10 milligrams, and I combine it with half of what you see there-- 250 cc's of sterile irrigation fluid. It's drawn out next to a sink-- which was an early lesson we learned-- and then we use the power wash connected to the scope. So I go in, I do all my cleaning as I'm going, the patient needs to be clean and in good control from their inflammation. And when I get to the cecum, they change out the bottle for me and connect it to the power wash. I then spray preferentially at the top of the colon-- and I mean top in terms of where gravity is-- so that I have to use less dye to cover the colon. It just drips down the walls, and it gets absorbed very beautifully. Here's an example of how chromo might work in a patient of mine. You can see that I'm coming back during my surveillance, and there's an obvious raised lesion, there. So you might say, well, why do you need chromo for that? Well, in this particular high-risk patient, when I sprayed the blue dye, this is a really nice example of how the absorption of the dye then highlights much more than just the raised lesion, it shows you the margins of it-- as you see as I pull back a little bit here-- and the pit patterns, here. That cerebriform-- it almost looks like a brain that you're seeing there-- tells us that that's most likely a dysplastic lesion, and that tells us what we need to do. So such in vivo pathology analysis by an experienced endoscopist or surgeon, can certainly be performed, and you have a better idea when you should be spending time removing things versus not. Here's another example. It's still images of a lesion I saw that was clearly there, but you can see when I spray it with methylene blue how much it became much more highlighted, including the margins. Which enabled me to subsequently perform an endoscopic mucosa resection. And then-- although this isn't usually necessary-- in this case, because it was a deeper lesion and I was in the right colon, I used some clips to close it. Now you can learn about pit patterns. This is just like anything else in medicine, it's experience and volume, and that's pattern recognition. Those on the left are more non-neoplastic and what you might see with hyperplasia. And those on the right are the more insidious and worrisome lesions that we look for. So over time, you really get pretty good at doing this. There remain a lot of challenges. Gastroenterologists are not running to the store to buy their methylene blue right now, because, in fact, they don't know how to do it. But also because they're worried it takes longer to do this. I would argue that when you use the power wash, you leave your instrument port open, so you can be doing biopsies and things move pretty fast. But we certainly understand the limitations. And, in fact, the evidence doesn't necessarily say this should be our preferred strategy in most patients. Now I say that, as my opinion, people are trying to move the field that way, and I'm not sure yet that we're ready for it. So you're not to leave this meeting thinking that you should all start doing chromo. I think that for specific patients, it's certainly appropriate. And for the patient who's had previous dysplasia it might be appropriate. And for patients who have PSC it might be an appropriate strategy, because they're at the highest risk. There was an international consensus meeting in San Francisco in March of this year, that actually did a systematic review. And then the board of individuals who were present-- including surgeons, pathologists, gastroenterologists, and patients-- voted on the strength of the evidence, and what the recommendations might be. And what they came up with were a few different summary statements. And these are not the official summary statements, this is my notes from the meeting, because we haven't published it yet. Number one-- high-def scopes are better than standard-def scopes for looking for dysplasia. Secondly, narrow-band imaging, or flipping the switch on your processor, is not better than white light, especially when you're using high-def scopes. So, although I use that routinely, it isn't necessarily going to make it into our guidelines. Chromo is better than white light, but is not necessarily for all patients. The newer technologies can enable ongoing surveillance instead of colectomy for everyone, especially for certain patient types. And a distinction is made between follow-up of visible and resectable dysplasia, versus that that might be found when you do a random biopsy, or that you didn't notice when you were looking for it. So the last thing I'll end with, is how do we incorporate all these advances to a smarter surveillance strategy, which is a better use of our resources? So how might we do this? Well, the first thing is to just keep in mind that you can stratify based on cumulative or combined risk factors. So the patient and disease-related factors that I've already gone over-- like PSC, family history, duration, degree of inflammation over time, and even on a single exam, males versus females, and your patients willingness to show up and follow up with you. Versus the dysplasia-related factors, like high-grade versus low-grade versus indefinite. The morphology-- flat versus polypoid, or invisible vs raised. And the field of factor synchronicity-- in other words, is there multifocal dysplasia versus unifocal dysplasia? As well as follow-up in multiple exams that have dysplasia versus a single exam. Now, the British Society of Gastro in 2010 embraced a lot of this-- without good evidence-- but tried to stratify their resource allocation. They said that after 10 years, you would start your exams and you would stratify based on low risk, intermediate risk, and high risk. And they recommended chromoscopy with dye spray for most patients. And based on your findings, you could go as far as five years between your colitis surveillance exams, which most of us are not comfortable with, in which there certainly is not evidence to support yet. On the other hand, it does make some sense to say that one way you can stratify your follow-up is going to be based on how well controlled the inflammation is, and what you find when you're on those exams. They also go on to say that there are some other considerations, as I've already mentioned to you. So it's not unreasonable to say that that patient who's always in complete remission, who's colon is clean, where you do a good examine. And even if you use chromo, and you find nothing, that, in fact, give them two years, or three years, before you bring them back again. Versus that one who is always inflamed, and who has PSC, who you're clearly are going to do it annually. So I'll end with a summary slide that says that, in fact, we are making progress in this field. I think that a smart approach to the patients at risk for dysplasia who have IBD is reasonable. And by defining those at-risk patients, and including inflammation as an important independent risk factor, you could actually decide how to best use your resources, and when it's appropriate to refer to your surgeons. My last point that I always make to my colleagues is that any dysplasia should really prompt a discussion about colectomy. Whether or not you go for it is a different discussion. But I often send them to our colleagues in the surgical department for an opinion, so at lest we're documenting, and the patient understands all their options. Thank you very much.
