Ryan Ungaro, MD, on Perturbances in Metabolic Pathways in IBD
Dr Raymond Cross discusses with Dr Ryan Ungaro how the Mount Sinai Crohn's and Colitis Registry, a prospective observational cohort, is being studied to determine if its detailed clinical information and omics data collected can identify new prognostic markers for disease progression in IBD.
Raymond Cross, MD, is director of the IBD Center at Mercy Medical Center in Baltimore, Maryland, and professor of medicine at the University of Maryland. Ryan Ungaro, MD, is an Associate Professor of Medicine in the Division of Gastroenterology at the Icahn School of Medicine at Mount Sinai in New York, New York.
TRANSCRIPT:
Welcome everyone to IBD Drive time. I'm Raymond Cross from Mercy Medical Center in Baltimore, and I'm delighted that my friend and colleague Ryan Ngaro from Mount Sinai here to talk about a recent research paper he published in Journal of Crohn's and Colitis entitled "Distinct Perturbances in Metabolic Pathways Associated With Disease Progression in IBD." Ryan, wlcome to IBD Drive Time.
Dr Ungaro:
Great. Thanks so much for having me, Ray. It's great to be here with you. And we usually will be talking some baseball, but today we're going to talk mostly IBD.
Dr Cross::
Correct. We can always talk about baseball for your fun fact that we can do.
Dr Ungaro:
That's true.
Dr Cross::
So before we specifically talk about your article as far as trying to stratify the new ACG guidelines have come out for both ulcerative colitis and Crohn's, and they do talk about assessing severity, which is prognosis as well as activity, which is current symptomatic and inflammatory activity. So what current tools do we have right now which can predict which patients are going to have an aggressive disease course in Crohn's or colitis or conversely, do we have any that tell us who the mild patients are?
Dr Ungaro:
Yeah, I think it is a great question and I think it's something that we encounter every day in our clinic trying to predict the future. In IBD is still a little bit of a waiting for Godot type of situation, particularly when it comes to biomarkers trying to find blood or tissue or stool markers that can really predict who's going to have an aggressive disease course or a mild disease course. And as you highlighted, there's this real disconnect that I think everyone's aware of between symptoms and prognosis, the idea of disease activity and disease severity; activity’s how this patient's feeling at this moment in time and severity is trying to be more of a picture of the totality of the disease course including prognostication. And right now, really there's primarily in a day-to-day practice for clinicians is clinical features and clinical features, generally speaking past predicting the future.
So these are things that were initially proposed in the AGA care pathway have been highlighted again in ACG clinical guidelines as well that really certain features in Crohn's patients and ulcerative colitis patients increase the risk of someone having an aggressive disease course. What we mean by that is someone who's going to need surgery, have frequent hospitalizations, depending on which studies you're looking at some definitions include escalation to needing a biologic, which some would say that might be standard of care for Crohn's disease. But in Crohn's, the features that have been consistently associated with a higher risk for disease progression are young age of diagnosis—generally speaking less than age 18—having extensive bowel involved and so extensive small bowel disease, multiple segments of disease involved. The definitions of that has varied widely, I would say at least more than 20 centimeters of small bowel disease and or colonic involvement as well as ileal involvement; having a history of stricturing or penetrating complications increases the risk as well; presence of perianal disease; and really those features have been consistently associated in epidemiological studies. Other things that have been associated have been the presence of deep ulcerations on endoscopy, ileal disease location, I think we all have patients where ileum is a frequent area of either fistulization and/or stricturing disease—why that is the case, I think there's still ongoing research there, but ileal location, and there's been some work on biomarkers in Crohn's disease, and this was things such as NOD2 as a genetic mutation, although that is likely more of a surrogate for ileal disease location, but antimicrobial antibody titers. So when I say that, I mean things like the ASCAs, antiflagellin antibodies, things that are antibody responses to something in the gut, either a yeast or a bacteria, that higher titers of those have been associated with increased risk of disease progression in Crohn's disease.
And those biomarkers have actually been incorporated into a tool that was developed actually by Corey Siegel called the Prospect Tool, and now is actually available clinically, where they combine certain clinical features together like disease location, as I mentioned, age, with some of these titers for antimicrobial antibodies to try to develop a risk profile for a patient of how likely it is they're going to have disease progression. In general, it's a nice tool to have, I think particularly if you're uncertain about a patient's risk profile or the patient is having a tough time making a decision about going on to an advanced therapy. Although the performance is still relatively modest, it is decent, AUCs in the range of 0.7, 0.75, which is okay, but still not fantastic. And we've done also some research trying to look at just these clinical features alone, the things I mentioned before, young age diagnosis, having extensive anatomic involvement, having a history of fistulas strictures or perianal disease. And really those alone actually perform fairly well, that if you have at least 2 of those clinical features that are high risk in Crohn's disease, that patients are 2 to 3 times more likely to have a complicated disease course over time, particularly in newly diagnosed patients.
So it's imperfect in Crohn's disease, but a bit of it is trying to put this piece together, these high risk features that you can assess clinically day-to-day in the clinic with the everyday risk benefit discussions we have with our patients about when they need to escalate therapy.
I think conversely, ulcerative colitis is a bit more unpredictable. We have similar types of guidelines from AGA and ACG about what are the high risk features—extensive disease, having deep ulcerations, having a high inflammatory burden, high CRP, low albumin, et cetera. But these are not necessarily reliable indicators over time in predictive models that you can predict who's going to need a colectomy in UC or really have an aggressive disease course. And that is something that is still, as much as there's still a need for improved prognostication in Crohn's disease, there's really limited data there right now in ulcerative colitis.
One emerging biomarker in ulcerative colitis that is our group at Mount Sinai led by Alexandra Livanos, has looked at a novel anti-integrin antibody, alpha V beta 6, and that those titers of that antibody seem to be associated with a more aggressive disease course in ulcerative colitis. And there's a couple of abstracts at DW suggesting that as well.
So all of this is to say is that it still comes down to, I think the clinical features and a little bit of the patient's history predicts their future. And to me with prognostication in IBD, we're still looking for blood biomarkers or tissue biomarkers, et cetera, to allow us to tailor therapy in terms of matching the patient's prognosis with a more advanced therapy in an appropriate fashion.
Dr Cross:
I just want to follow up. I agree with everything you said, Ryan. I just want to follow up with 2 things. One is I think for UC, if you forget all the clinical factors that we talk about, to me, I think the one critical one is if you've been hospitalized for your ulcerative colitis. And I think to me that's a big red flag that this patient needs an advanced therapy. You were sick enough to get into the hospital. And I think sometimes when we think about disease extent, sometimes that can actually be a bias the other way— that if you have only left-sided disease or only proctitis, this idea that you're not going to go to an advanced therapy not involving the whole colon is actually incorrect. But I think for me, that's a big one. And then this anti-integrin antibody, the other really cool thing about it is this concept of interception, where you potentially could identify a high risk patient and assuming we have some intervention that we could do, that you might be able to even prevent or find IBD very early when it's very mild and treat it more effectively. If I remember right from those studies up to a decade before, you can see the emergence of this antibody.
Dr Ungaro:
Yeah, absolutely right. And that's another topic that's near and dear to my heart and a lot of us in Mount Sinai, is that the idea of disease prediction and prevention. Can you identify people at risk of developing disease to either closely monitor them or intervene to delay the onset or prevent the onset? And you're exactly right that in Crohn's disease, actually, to just draw a quick contrast. There's actually many markers, proteomic markers, metabolomic markers, the ASCAs that I mentioned before, the antiflagellin that you can see elevated years before someone's ever diagnosed with Crohn's disease. And for a long time there was really almost nothing that we could see in the prediagnostic phase for UC. And this anti-integrin is really one of the only markers that seems to actually be present years before someone's diagnosed, indicating some sort of predisposition to developing disease.
And then just to quickly piggyback on something you mentioned about UC, I totally agree with what you're saying that especially the acute severe UC patient, I feel like sometimes there's a bit of a natural history event that it's portending a more aggressive poor prognosis. Obviously there's differences in thresholds to hospitalize somebody based on your hospital and ER clinical setting, but that is a big one.
And then you mentioned disease extent, and actually that's something that I think there's some research on this now, too, that some of your sickest, hardest to control patients that need to get escalated to therapy quickly are those patients who go from limited disease, whether it's proctitis, left-sided colitis, to pan colitis. So that disease extension event portends a very poor prognosis. And those are patients that if you're seeing, they're typically very symptomatic, they may end up hospitalized. And those are patients that definitely need escalation of therapy rapidly. So yeah, those are 2 scenarios you pointed out that are definitely good prognostic indicators for UC at least.
Dr Cross:
So let's talk about your study. So first, why don't you explain to the listeners what the study design was?
Dr Ungaro:
Sure. So this was a study where we utilized a prospective observational cohort for Mount Sinai called the Mount Sinai Crohn's and Colitis Registry, or M-S-C-C-R for short. And essentially this was a large effort to enroll hundreds of patients with Crohn's and colitis in a registry where they had multiple biosamples taken at the time of a colonoscopy. So these patients had lots of blood samples taken. They also had biopsies, taken detailed clinical information at a single time point, and then they were followed through their EHR. So they gave consent to have prospective follow up through the EHR. They also were allowed to be recontacted, although we haven't really utilized that fully to be totally honest to full extent, but it is a rich registry of molecular data where these patients had from those samples already run different assays. So things such as genetics, obviously they had metabolomics, they had RNA sequencing done on biopsies and on blood as well.
And so what we did was ask the question—we have this cohort now that it was originally collected 5-plus years ago. Now they've accrued follow-up time through the EHR through their medical records. Can we look back at that baseline visit where they had all this detailed clinical information and omics data collected, and can we help to understand if there's some new prognostic markers for disease progression? And so the study design was, as I mentioned, a prospective cohort where in this specific case we want to understand if metabolomics or metabolites, circulating metabolites in the blood, could potentially be predictors of a complicated disease course, which in this case we defined as a composite outcome of any new surgery, new complication, or need to change therapy, whether that is needing to go to a new biologic or escalate to a biologic or needing steroids. It was a composite endpoint. And then patients were followed for a median of 2 to 3 years, and we looked at that baseline metabolic profile and clinical features to see if those were at all associated with disease progression.
Dr Cross:
So obviously important outcomes. And what were the key results, Ryan?
Dr Ungaro:
So we used a platform called Metabolon. For those who aren't familiar with it, it's a commercially available platform that gives you a very comprehensive panel of circulating metabolites. And for those who are familiar, the metabolome, metabolites, these are all small molecules that really are reflective of multiple inputs: so host genetics, microbiota diet, lifestyle inputs, that all then result in your body's metabolic profile. So these are things that are like nucleotides, carbohydrates, amino acids, lipids, et cetera. And what we found was that in both Crohn's disease and ulcerative colitis, we saw multiple metabolites that were either positively associated, so increasing the risk of having a complicated disease course, and also interestingly, other metabolites that were associated with decreasing the risk of having a complicated disease course. And with all of these types of studies, you want to understand for any kind of prognostic study, you're looking at biomarkers, I think it's key to understand what's the underlying clinical features, and if the clinical features are as predictive or once you adjust for clinical features, if the biomarker results are negated, then it's not really that valuable. And so when we looked at adjusting for multiple clinical variables, so we looked at endoscopic scores like SES-CD, Mayo score, we looked at history of surgeries, the medications patients were on, et cetera. Even when you adjusted for clinical features, you still had multiple metabolites that were associated with increased or decreased risk of disease progression. And I think this just hints at the importance of metabolism and metabolites as a new area for research to understand underlying IBD biology and potential new types of biomarkers to help us rest stratify patients in the clinic
Dr Cross:
Strengths, it's obviously a prospective cohort design. So what are the strengths and weaknesses of the study, Ryan?
Dr Ungaro:
I think the strengths are that it was a prospective study; it was a large study, it was over 600 patients, so it had a good sample size that allowed for statistical adjustments and to look for statistically significant changes. I think the limitations obviously are that it was a single center study and we know that EHR studies are only as good as EHR, so junk in junk out. And so the data, just the risk for potential incomplete data that you may have someone who is getting care elsewhere and you might miss some events potentially. And then I think another limitation is that we did not have an external validation cohort, which is always important for any of these types of biomarker studies—that there's been a ton of biomarker studies that whenever you see in the literature, I think it's important to look for that external validation cohort and at least plans to do that.
And I think a prime example of that in the literature recently was the PROFILE trial where they saw in one cohort that there was a blood profile that was associated with disease prognostication, but when they tested it in a clinical trial that it didn't pan out. So I think this is still an initial kind of preliminary, intriguing finding, an intriguing study that maybe highlights some new biomarkers of prognostication for Crohn's and UC. And I think I also, interestingly, I think the take home also that metabolome, metabolism, immunometabolism are all emerging areas in IBD in terms of understanding are these other underlying pathways that may be contributing to prognostication, disease progression, treatment response, et cetera. So I think it's interesting conceptually that this idea of metabolism and metabolism is something that is underappreciated in IBD and at least warrants further study.
Dr Cross:
I can think of at least one more strength and maybe one potential weakness. The strength being is these patients, unlike most of the things you do in the EMR, these patients are deeply phenotyped. So a lot of these confounding variables that we don't have, like when we do an Optum study, you can't really get the patient level variables that you need. You have that here. When you use steroids or an advanced therapy change or surgery as an outcome and a composite, there is some subjectivity to that. I can't imagine how, if there was bias, I would think it'd be nondifferential bias. I can't imagine how that would be impacted by someone with a certain profile, because you weren't even aware of what that profile was.
Dr Ungaro:
Right. I think you would suspect that you may see, you'd be more concerned if maybe there was not any effect, that you're missing surgeries, hospitalizations, et cetera, outside of the health system. But yeah, you're right. I think at the time of ascertaining, you don't know how the metabolic profiles are influencing that. There shouldn't be any ascertain bias or things like that either.
Dr Cross:
So I think the clinical implications I think are pretty obvious. I mean, if you have a patient at any time point that you're trying to risk stratify, if you can validate this and make it a simple result that clinicians can interpret, that's obvious. I was just thinking about Anita Afzali and I just did a Drive Time about the new Crohn's guidelines and thinking about the treat to target concept and what do you do when someone isn't healed and could you use something like this to turn that into a high risk or a low risk patient? Or how about your intermediate risk for post-op recurrence where you're not quite sure if you need to start a therapy? So I think the utility of this is sort of endless, if you think about it.
Dr Ungaro:
Yeah, no, and I think the nice thing about metabolites is that they're actually things that have been used in clinical practice actually fairly readily in both IBD and other, even more so in other fields. So in IBD, we look at thiopurine metabolites, for example. And so they're easier, relatively easily to implement in a clinical assay. Other clinical fields, other disciplines use these all the time. Cholesterol levels are metabolites. You can look at glucose levels as metabolites. So there's a lot of different examples of that. Metabolism and metabolites can be implemented clinically as monitoring tools. And I'm curious to see also, I think as follow up to this, do these markers tell you about certain pathways that could be targeted in people who are treatment refractory? There's a lot of interest in adding on GLP. There's ongoing trials that are emerging now to add on new GLP receptor antagonists, that these are actually maybe potentially also highlighting pathways that can be complimentary to some of the immune pathways we’re blocking as add-on therapies or combo therapies. And that's obviously a bit of a stretch from this study, but that's just I think just in general thinking about metabolism and immunometabolism and how it could impact IPD
Dr Cross:
And Ryan, next steps for this, I presume you're going to try to validate this in another population.
Dr Ungaro:
So we want to do this in different cohorts, more than one cohort. And our next push is to try to do this in 2 scenarios. One, something you mentioned already is in the preclinical scenario. So we have access to unique cohorts of patients who we have blood on from years before diagnosis, United States military cohort, the PREDICTS cohort is the prime example we use in a lot of research at our institution. And the second would be inception cohorts of patients that are recently diagnosed. Can this potentially be helpful in stratifying, phenotyping patients? And maybe we'll be talking about, we have the Montreal classification, maybe we're thinking about metabolotypes for patients as well in the future if this continues to grow.
Dr Cross:
I just want to remind the listeners that IBD Drive Time is sponsored by the AIBD network. We're available on Spotify and Apple Podcast. Our next regional Advances in IBD is going to be in Cleveland October 11th to October 12th.
Ryan, before we go, what's your fun facts? So tell me something about yourself that I don't know, or maybe that the listeners don't know.
Dr Ungaro:
Sure. So I think I'll give you a fun fact is my, for many, many years I was a trombonist. So I played trombone growing up and through college and med school, and I was in jazz orchestra in particular. So I did a lot of jazz trombone playing for most of my younger years, and I haven't picked it up in a while, but something I've been inkling to do. But that was in jazz orchestra and then also in high school was, this was in the nineties, so it was like rock ska bands as well. So I used to be a musician in another life.
Dr Cross:
Very cool. And you played a cool instrument. So I think that given the number of these I've done, if I'm guesstimating, one out of every three guests mentions being a musician at some point in their life. So we definitely have enough varied instruments and things that we could definitely have an IBD Drive Time band for sure.
Dr Ungaro:
It's a great idea. We should crack out the old instruments. That'd be great.
Dr Cross:
Sounds good. Ryan, this has been wonderful. Thanks so much for doing this, and we hope to have you back soon.
Dr Ungaro:
Great. Thanks so much.
