IBD Drive Time: AGA Guidelines for Biomarkers in Ulcerative Colitis

In this episode of IBD Drive Time, cohost Raymond Cross, MD, talks with Siddharth Singh, MD, on the new guidelines from the American Gastroenterological Association on caring for patients with ulcerative colitis.

Raymond Cross, MD, is professor of medicine and director of the IBD Program at the University of Maryland School of Medicine in Baltimore. Millie Long, MD, is a professor of medicine, vice chief of education and director of the fellowship program in the Division of Gastroenterology and Hepatology at the University of North Carolina at Chapel Hill. Siddharth Singh, MD, is an associate professor of Clinical Medicine at the University of California San Diego in La Jolla, California.

TRANSCRIPT:

Welcome everyone. I'm Raymond Cross from University of Maryland School of Medicine, and I'm one of the cohosts for IBD Drive Time. I'm delighted to have my friend and colleague, Sid Singh here today to talk to us about the new AGA guidelines on biomarkers in UC. Sid, welcome to IBD Drive Time.

Dr. Singh:

Thank you so much for having me and thank you for the opportunity to share this with all of you.

Dr Cross:

Wonderful. I just quoted the guidelines today with one of the trainees, so maybe start with telling the listeners what the rationale for the new guidelines in UC was, but what was the rationale by the AGA for these new guidelines? Why do we need these?

Dr. Singh:

Yeah, no, absolutely. I think there is an increasing recognition that symptoms are an imperfect marker of disease activity for patients with ulcerative colitis. And we've gradually moved towards a combination of symptoms and endoscopic assessment in a lot of our patients. At the same time, endoscopy is an invasive test that's expensive, and quite a few of our patients may not want to do them repeatedly. So biomarkers fit in very nicely in the middle wherein they can provide some objective evidence on the degree of inflammation that may be present in patients with ulcerative colitis, those with or without symptoms. They're frequently used in clinical practice and there's a lot of practice variability as to how people use them, how people respond to them, what cutoffs they use.

Our goal here with the guidelines was to inform the best use of these biomarkers based on available evidence and to see how these should be used, what cutoffs, and how do they correlate with outcomes. The rationale here was to inform better use and a more standardization of the use of biomarkers when we are monitoring our patients with established ulcerative colitis.

Dr Cross:

Wonderful. And I'm going to put you on the spot here. I think this is sort of a trick question, but what's the best biomarker for UC? Was your group able to determine what had the best performance as a biomarker?

Dr. Singh:

No, that's an excellent question. We focus primarily on established biomarkers, which are commonly available. So I will probably not comment on any proprietary biomarker or a biomarker that has shown great performance in a research setting. But we focused on 3 biomarkers, and that was fecal calprotectin, fecal lactoferrin, and serum C-reactive protein.

The biggest body of evidence was for fecal calprotectin, and that was the one that we were able to look at across multitude of studies, examine different cutoffs, how different cutoffs perform. Even though serum C-reactive protein is so commonly ordered with every blood panel that we do, surprisingly, there was positive data, especially looking at different cutoffs for CRP for its performance. And fecal lactoferrin, similarly, there was some selection in terms of the number of studies that were published using fecal lactoferrin. So I would say the biggest body of evidence, and perhaps by extension, the best biomarker for which we have most confidence is fecal calprotectin.

Ray

Yeah, I think this is more relevant in Crohn's disease where I tell the trainees that there is no perfect biomarker, that you don't just order a CRP, a sed rate, and a calprotectin on every patient. You really want to look back and see at the time that Jane Doe was having symptoms, what biomarkers were checked and were they reflective of the endoscopic activity or the imaging activity that was seen. But I do think for ulcerative colitis, the calprotectin is pretty reliable. I think maybe I have a handful of patients where it doesn't help me at all, but generally I think it's pretty good. Any subtypes of UC where calprotectin is less useful? So proctitis versus left sided versus extensive or pancolitis?

Dr. Singh:

Yeah. First of all, I want to completely echo what you said. I think anchoring the biomarker with an endoscopic activity for a patient is probably the best way to figure out the performance. So when somebody's flaring up, anchor and see which biomarker correlates best. And the same thing when that same person goes into remission, anchor and see whether the biomarker comes down, and I think that provides you the best biomarker for that particular patient.

We try to look at the performance of calprotectin in different subsets by disease extent. And while there is data to suggest that perhaps performs better in people with left-sided or more importantly extensive colitis, the correlation with endoscopic inflammation may be less strong for proctitis. So we were not able to come up with separate thresholds just because the data was just not that nimble to allow us to say for proctitis use different thresholds. So we basically said, to your point previously, let's take a look whether even in proctitis patients does calprotectin correlate with endoscopic activity or not.

Dr Cross:

I think for the listeners that haven't been on one of these guideline committees, I was fortunate enough that I was on the guidelines for mild to moderate UC. Can you explain to them simply how you grade the evidence and how you determine your recommendations to give them an idea when they look at ... because some of the recommendations were conditional with low or very low certainty, so if you could quickly explain that, I think it would be helpful.

Dr. Singh:

No, absolutely. So again, it is a rigorous process, multistep process. Again, all the recommendations initially start off with developing a focus question as to what we are trying to answer. And all recommendations are meant to be actionable so that you can read the recommendation, apply it in practice. Start off with a systematic review that's going to be answering that question, boil it down to a meta-analysis if need be, and then we look at the risk of bias in the individual studies as well as use what we call a grade framework to assert the confidence in those estimates.

This guideline for biomarkers is a little more challenging because while for interventions you can grade the evidence and say, treatment A works well for disease, and you can say with what confidence you think it works, how well. For biomarkers, we were now trying to infer on the performance of biomarkers for a downstream effect. So if the biomarker is high, what would it mean in terms of presence or absence of inflammation and you acting on that inflammation with treatment adjustments, changes. So there was a lot of indirectness in going from step one of what is the sensitivity specificity of a test into looking at it in different clinical scenarios where the pretest probabilities will differ. And then looking at what is the likelihood that I'll make a wrong decision if I rely on this combination of symptoms and biomarkers.

And so the panel set thresholds and said, we don't want to be wrong, false negative rates of less than 5%, false positive rates of less than 5%. We were rigorous in how we implemented these biomarkers in combining these clinical scenarios in which these are applied.

Dr Cross:

And so I was going to ask you a follow-up, but maybe I'll summarize it and then ask you a follow-up. So one of the things that I thought was neat in the guidelines was that your group did a meta-analysis of the data with fecal calprotectin and showing that if you had a normal calprotectin, your risk of relapse was about 15%, whereas if the calprotectin was elevated, it was about 65%. I might be off a little bit there, but I think the listeners that are saying, well, why would I be checking this in a patient who feels well? Well that's why, because they're more likely to relapse. Is there data that intervening in those patients can prevent a relapse based just on the biomarker?

Dr. Singh:

That's an excellent question. Yeah. So we initially thought, let's look at randomized controlled trials ideally, where people with Crohn's and UC are otherwise asymptomatic. We check a biomarker. And an ideal way to answer this question is to actually do a clinical trial where you're checking a biomarker as part of a biomarker based monitoring strategy and acting on it to see what the outcomes were.

There was one small, randomized trial of 50-odd patients with mild ulcerative colitis who were in symptomatic remission and treatment was adjusted if the calprotectin was more than 50, dose of mesalamine was increased, and the other group was randomized to no treatment changes. And within that small clinical trial, they observed at the end of 48 weeks that the proportion of patients who had a relapse was higher in where you did not make any treatment adjustments, but in other ways the risk of relapse was lower if you increase the dose in response to an elevated calprotectin in an asymptomatic patient. We clearly do need more of these randomized trials to better inform how biomarkers are useful in the practice.

Dr Cross:

I don't remember if you were part of the MARQUIS study, but we were, and the MARQUIS study that Mark Osterman led showed that you could adjust mesalamine dosing to normalize the calprotectin. It didn't show the difference in relapse rates, but it was about 100 patients or something like that. And it did at least show that you could change the value, for what that's worth.

Dr. Singh:

Yeah, exactly.

Dr Cross:

So I just started a patient today on upadacitinib. I'm going to bring that patient back in about 6 to 8 weeks to assess response. Patients feeling well, frequencies normal, not bleeding. What are the thresholds for the different biomarkers you mentioned that I should be looking for? Although acknowledging maybe they're not going to be perfect at week 6 or week 8, but what am I eventually shooting for maybe at the midpoint like week 10 or 12?

Dr. Singh:

Yeah, no, absolutely. That's an excellent question. I'll answer this question in 2 ways. One of those would be the scenario in which you are presenting and another in terms of the thresholds, because the thresholds are a little bit different. So when we looked at the data, again, thresholds could be looked at best for fecal calprotectin, but the other ones were just elevated or normal just because the data wasn't present for different thresholds. But for somebody who is symptomatic, you make a treatment adjustment by week 10 or 12. If you are able to achieve a fecal calprotectin of less than 50, then you can be reasonably confident that that patient is in endoscopic remission.

However, if you were to look at a different patient and let's say you're doing a follow-up during maintenance, and this is closer to 20, 26 weeks, then we failed the performance of a fecal calprotectin threshold of 150 would probably be best. I do want to add here that again, like you said, Ray, 150 is not a clean number in the sense 151 is bad, but 149 is okay. It's never like that. The way we parse out the data, we actually said 150 plus minus 50 for performance. So in effect, a range of a hundred to 200 may be okay for a routine patient during a routine follow-up for calprotectin thresholds such that if the levels are above 150, or if you want to be purist then above 200, then that may imply a presence of significant inflammation.

Dr Cross:

And normalization of lactoferrin and CRP? Assuming they were abnormal at baseline.

Dr. Singh:

That is correct. General thresholds for CRP were 5 milligrams per liter. For lactoferrin we usually relied on whatever the manufacturer said, and that threshold is usually 7.25 micrograms per gram of stool.

Dr Cross:

Great. Before I ask Sid a couple more questions, I just want to remind the listeners that IBD Drive Time is now available on Spotify, Apple Podcasts, the Gastroenterology Learning Network and SoundCloud, and we are sponsored by Advances in IBD and the Gastroenterology Learning Network. And speaking of Advances in IBD, the next live regional is taking place July 20 and 21. Millie and I will both be speaking, so there's still time to register for that excellent virtual course.

So Sid, I think we've been in this treat to target concept for a while, particularly for Crohn's, and I think many providers are getting more used to doing some type of endoscopy or imaging around 6 months after starting or changing therapy. How should they think about this in place of endoscopy, and are there subgroups of patients that even if the biomarkers look good, that you would be inclined to still do that assessment?

Dr. Singh:

Yeah, that's an excellent question. I do want to clearly state, we did not intend to modify the treat-to-target paradigm in this, and we were not in a position to answer that question for reasons like you mentioned, we do not have clinical trials comparing biomarker based monitoring versus endoscopy based monitoring, and that was a knowledge gap. In our practice, and the way that we want to emphasize in the guidelines is we still do a treat-to-target such that if you're able to confirm achievement of endoscopic healing six months after any treatment adjustment, that would still be the main way of managing most of your patients.

Biomarkers can be useful so that, let's say a patient has achieved endoscopic healing, how do you monitor the patient following that? How do you make sure that over the next two or three years, you're not repeating the scope every year to monitor that patient? And similarly, Ray, to the patient that you alluded to who was started and recently adjusted on a medication, biomarkers can be used to make interim changes, but I think the end goal here is still going to be endoscopic healing.

The reason for that is our biomarkers were studied for looking at an outcome of ruling out moderate to severe inflammation, so Mayo endoscopy score of 2 or 3. The biomarkers are not very nimble at looking at Mayo 0 versus Mayo 1. And again, as the field has evolved into focusing more on Mayo 0, we may not be able to get to that answer just with biomarkers. So I do want to emphasize that for most patients, the treat-to-target paradigm would still involve checking and confirming an endoscopic remission or endoscopic healing endpoint 6, 9 months after your treatment initiation and adjustment.

Dr Cross:

Yeah, I think for Crohn's disease, I'm pretty good about rescoping them after a change or start, although it may not always be at 6 months. In someone who I'm just making minor adjustments, I'll give them a little bit more time just to be patient-centric and recognizing that an invasive procedure can be tough to keep doing those every 6 months.

I do think that there's a much tighter connection between symptoms and endoscopy with UC. So I'm being honest, if I have a patient who normalizes their bowel movement pattern, who's not seeing any blood, and I've got a biomarker that was abnormal at the start and is now normalized, I may not scope that patient unless they need dysplasia surveillance or maybe they're a visual person that needs the pictures. I do offer it to them, but I feel pretty comfortable if the cal pro or whatever biomarker is normal by not scoping.

But I think your point is well taken that at least for that first assessment after starting therapy, you can often tie it with a biomarker and you do it, and then if you're going to make an adjustment in 6 months, maybe you don't need to do a second scope on that patient. You can just do the first scope. So I think that's a very reasonable approach, and it may keep our division chiefs and chairs happy because we do generate some revenue with these procedures. But also, these biomarker guidelines were designed to potentially reduce cost as well.

So in that patient scenario where you're starting a therapy, you're checking your biomarkers, now you're going to do a sigmoidoscopy or colonoscopy. The patient feels well, how do you determine whether you're going to make a change if the endoscopy doesn't look as good as you want it to look? Let's say there's Mayo 2 disease involving a third to a half of the colon, so not just a little strip of proctitis left. How do you determine when you're going to abandon something in someone who feels well?

Dr. Singh:

That's an excellent question, and that is actually a key knowledge gap for us. The treat-to-target paradigm would suggest that you need to keep escalating till you achieve your target that you and your patient have set. And if that target was set as endoscopic remission, then you should keep adjusting. In my practice, I'm typically comfortable with adjusting their index medication in the sense if they're on a standard dose of infliximab, I would check a drug level. If the drug level is on the lower side, I would escalate infliximab because I think that's a low-risk treatment adjustment from a patient perspective, risk of side effects is low. The cost is going to be higher, but it's unlikely that I'm going to fall back.

Whether I switch somebody out to a different mechanism of action or a different class altogether is an answer that I do not know. And this is where I put in a plug in where actually there's a PCORI funded clinical trial that we are doing called the COSENT trial, which is exactly focusing on these patients who are asymptomatic, optimized on their medication, but you do a scope or do an imaging study, they have significant inflammation ongoing. Should you switch them to a new medication or do you just let them be on their existing medication? This may be the best that we get with the medication, and that may be good enough to keep a patient in reasonable disease control.

Dr Cross:

Yeah, I completely agree. I try to squeeze every last bit of toothpaste out of the tube if I can. And we know that that approach generally is not harmful for our patients other than there is a financial aspect of that for the payer. I think where I'm more comfortable switching is, and this is I think more relevant in Crohn's ,when you still have fairly deep ulcerations. I think we know the natural history there, of course, the postoperative population because they don't sense their Crohn's until it's too late. So I think if it looks bad and we're not just talking aphthous ulcers, but you're getting deep ulcers and you've maximized, then I'm more likely to change. I don't think that's really as applicable in ulcerative colitis in general. And it also depends on your treatment journey, right?

Dr. Singh:

Yes.

Dr Cross:

If you have nothing left, I would argue, why are you looking, unless it's for dysplasia surveillance? Because if they're not willing to have a colectomy, you should not order the test because you're not going to use that information.

Dr. Singh:

That is a very important practical point. I think having that discussion upfront with your patient as to what are the changes that we would potentially be making depending on the results of the test. I think that applies to biomarkers when you're testing them. That applies to a scope when you're treating patients. I think that is a critical component here, and we try to emphasize in the guidelines is make sure you know what is your next treatment step, and if you and the patient are on board with that treatment step, that's the best time to use your test. Otherwise, you just have information and a cognitive dissonance between the test results, the clinical scenario, and where your patient is.

Dr Cross:

Yeah. And you and I have those patients that say, "I feel so good on this medication that there's no way I'm changing unless I'm starting to have symptoms." And you have to respect that, where people came from going 15 times a day and now they're going three times a day with no blood, off steroids and they're happy. So when they say that perfect can sometimes be the evil of good. So that is something that you and I think about all the time. Okay. Fun question. Sid, tell the audience, and maybe even me, something about yourself that we may not know. So this is what I call the fun fact.

Dr. Singh:

I can be up at any time of the night to watch cricket, and unfortunately, since most cricket happens in India or outside of the US, it's often 3:00 in the morning that I'm up and watching cricket. So I may not be the most functional person if there's a World Cup or some important cricket series going on.

Dr Cross:

So how do you manage that? You go to sleep at 9:00 and you get up and then you start your day at 3:00? Or it depends?

Dr. Singh:

It just depends. Yeah, exactly. I'll just start my day early and I'll confess to myself that today's not going to be day that I'll be productive outside of the minimal responsibilities that I have at work.

Dr Cross:

But you never do that before your endoscopy day.

Dr. Singh:

Definitely not. Lots of coffee.

Dr Cross:

All right, Sid, this has been wonderful. It's timely. I think the listeners are going to really find this useful. Thanks again, and hopefully we can have you back on IBD Drive Time.

Dr. Singh:

Perfect. Thank you so much for this opportunity.