IBD DriveTime: Manasi Agrawal, MD, on First-Line Resection for Crohn's Disease

In this episode of IBD DriveTime, Dr Manasi Agrawal discusses her research into surgery as a first-line therapy for Crohn's disease with host Dr Raymond Cross.

Raymond Cross, MD, is a professor of medicine and director of the IBD Program at the University of Maryland School of Medicine in Baltimore. Manasi Agrawal, MD, MS, is an assistant professor of bedicine at the Icahn School of Medicine at Mount Sinai and a gastroenterologist at the Mount Sinai Hospital, with focus in inflammatory bowel disease.

Any views and opinions expressed are those of the authors and or participants and do not necessarily reflect the views, policies, or positions of the AIBD network or HMP Global its employees and affiliates.

Dr Raymond Cross:

Welcome everyone to IBD Drive Time. I'm Raymond Cross from the University of Maryland School of Medicine and I'm delighted to have Manasi Agrawal from Mount Sinai here to talk about a recent research publication of hers in gastroenterology. Manasi, welcome to IBD Drive time.

Dr Agrawal:

Thank you very much Ray.

Dr Cross: (00:39):

So you recently published a paper in gastro on the difference in clinical outcomes between patients treated with an early ileocolic resection versus anti TNF therapy and specifically it was anti TNF, not other biologics. Can you just briefly describe the rationale for the study?

Dr Agrawal: (00:59):

Yes, absolutely. So there has been growing interest in early ileocecal resection as primary treatment for Crohn's disease and we learned a lot from the lyric trial, which was a randomized clinical trial to compare early ileocecal resection and infliximab treatment. And they found that one year outcomes were comparable between the two arms with respect to quality of life. And when they did a long-term retrospective analysis of these data, they also found very interesting results which were in favor of ileocecal resection, meaning supporting the role of ileocecal resection as primary therapy for limited Crohn's disease. However, real world data on this topic are rather limited and so we wanted to conduct this population-based study to truly compare the two arms and to understand if and how ileocecal resection may play a role as primary treatment for limited ileal and ileocecal Crohn's disease.

Dr Cross:

Wonderful. So for your study, what was the study design and what was the patient population?

Dr Agrawal:

Yes, so this was a nationwide population-based study, which we were able to do in collaboration with our wonderful colleagues in Copenhagen at the Predict Center, Christina Aileen, Tine Jess, Anthony, and others. And so for this study we used cross-linked Danish registers between a study period of January 1, 2003, to the end of 2018 and we identified a very specific group of Crohn's disease patients. We identified individuals who were diagnosed with Crohn's disease during the study period and who received either anti-TNF treatment or underwent ileocecal resection shortly after diagnosis—so either within 30 days of diagnosis to up to 1 year of diagnosis. We excluded individuals who had any evidence of perianal Crohn's disease because we estimated that a diagnosis of perianal CD warrants anti-TNF treatment or other biologic treatment, we excluded individuals who did not have adequate follow-up for us to be able to get enough variables on them both in terms of exposure variables and outcomes.

And finally we included only those individuals who had confirmed ileal or ileocecal disease based on data from the pathology registers. So for the purpose of these definitions, we combined cross-linked data from multiple different registers including their civil registration system, the pathology registers, like I said, the disease diagnosis codes as well as medications. And using this algorithm and these criteria, we finally included in our study 581 individuals who underwent early ileocecal resection and 698 individuals who received anti-TNF as the primary treatment. Our main purpose was to try to mimic the LIR!C trial as much as possible to be able to answer the specific question of early ICR versus anti-TNF for limited Crohn's disease.

Dr Cross:

And so with the 30-day criteria, you basically excluded people that presented a diagnosis with an obstruction or an appendicitis-like condition or intra-abdominal abscess, which is completely appropriate because they're going to be managed surgically. And if I remember right, I was actually at ECCO— where you did a beautiful job presenting this by the way, and Copenhagen speaking of beautiful is a beautiful city, so it was fun being there. But if I remember right, this was about 5% of the total population. Is that correct? Something like that.

Dr Agrawal:

So it was about 20% of the study population who had diagnosis codes associated with obstruction or stenosis or fistulizing disease, an internal fistula that is, and we actually included these individuals considering that this was a large proportion of individuals who underwent ileocecal resection because these complicated patients would shift the results towards the null and you are getting, if we have good results even after accounting for these sicker patients, then you would expect that an uncomplicated disease, the estimates would be more in favor of ileocecal resection.

Dr Cross:

Okay. So they could have complicated disease they just couldn't have had within 30 days of diagnosis and there was more patients of the total population treated with surgery or biologic within the first year than what I remember. Is that correct?

Dr Agrawal:

That is correct.

Dr Cross:

Okay. I'm glad I asked you because I remembered that incorrectly. So what were the primary outcomes as far as clinical outcomes you were looking at?

Dr Agrawal:

Yes. So the primary outcome we defined as a composite of the following adverse long-term outcomes. These included hospitalization for a Crohn's disease-related indication, systemic corticosteroid use, Crohn's disease-related surgery, and a diagnosis of perianal Crohn's disease. So any one of these outcomes constituted as a yes to the primary outcome. And then we analyzed each one of these outcomes separately as secondary outcomes.

Dr Cross:

And what were the high-level results?

Dr Agrawal:

The high-level results were that with respect to the primary outcome, meaning a composite of these adverse outcomes, the adjusted hazards of the composite outcome was 33% lower with ileocecal resection compared to anti-TNF therapy, meaning that individuals who underwent ileocecal resection did much better compared to anti-TNF therapy. And the variables that we adjusted for in this scenario included age at diagnosis, patient sex, year of treatment, number of hospital contacts for any indication, and number of unique prescription medications as a marker of comorbid conditions, systemic corticosteroid exposure, immunomodulator exposure, and all in the year prior to index treatment. With respect to the secondary outcomes, we found that systemic corticosteroid use as well as Crohn's disease-related surgery for these outcomes as well ileocecal resection patients did better compared to anti-TNF group. Whereas for Crohn's disease-related hospitalization, while the estimates were in favor of ileocecal resection, numerically the 95% confidence interval did cross 1 and similarly for perianal Crohn's disease.

And then in addition, we conducted some sensitivity analysis and stratified analysis to understand if there were any differences in these estimates between the 2 groups based on variables such as sex, age at CD diagnosis, year of treatment, and prior medication use. And we found that in each of these stratified analyses, for the most part, all of these estimates were very consistent, all in favor of ileocecal resection. The only exception was the pediatric age group of Crohn's disease diagnosis. However, our numbers were very small in this specific stratum and so it would be difficult to make an interpretation in that regard. But the P value for the trend test was nonsignificant in each of these analyses, meaning that there was no interaction across these additional variables. And in each of these subgroups, ileocecal resection appeared to be better than anti TNF therapy for long-term adverse outcomes.

Additionally, another important result that we thought was very relevant clinically was that we then did subanalysis to understand medication use and subsequent surgery in each of the two groups, ileocecal resection and anti-TNF. And we found that approximately 50% of individuals who underwent surgery were on no treatment at 5 years of follow-up. So this tells us that surgery may be a very good option for at least a subset of individuals and they may require no treatment following surgery on follow-up. Of course, we need to understand who these patients are and how we can identify the right treatment for the right patient.

Dr Cross:

Yeah, that's exactly what I was going to ask you about the follow-up, about that 50%. I wanted to test my memory again, see if I remembered it correctly. So lots of strengths using this type of cohort, but what were your just strengths and limitations of the study?

Dr Agrawal:

Yes. So some of the strengths of our study are the use of a population-based cohort, which makes it unselected. We had for over 15 years, which was very, very informative, we had complete or near complete follow-up. And we also have high accuracy of exposure and outcome variables, which we have looked into in previous studies. So all of these strengths of this study help towards the robustness of our findings and make it clinically relevant. Having said that, our study also has limitations. Of course, these are administrative data that we used secondarily for the purpose of research, we do have limited phenotype data, for example, smoking as well as disease progression-associated variables. And so that makes it a little bit harder to identify who is the right patient for ileocecal resection. And I think this sort of segues into future studies to address these unanswered questions. And of course with observational data there's a risk of unmeasured confounding. And so similarly in our study there is that limitation. Also,

Dr Cross:

I've heard people talk about that with the following the patients postoperatively, there's not information on endoscopy. So you don't know if disease is recurring and it's just a matter of time before they start treatment, but you did the best you can with the data that you have. And before I ask you about the clinical implications, I think we'll talk about that a bit.

I just want to remind the listeners that IBD Drive Time is sponsored by the AIBD Network and that we are available on Spotify and Apple Podcasts. So if you go to Gastroenterology Learning Network, you'll see IBD Drive Time and you can subscribe so you don't have to wait for the email to listen to the fun podcast.

So I'm sure you learned a ton doing this and using this dataset, but clinically, what's the message here? What's the message for the listeners and then for you and your clinical practice perhaps I hate 2-part questions, but I'm going to give you a 2-part question. How has it changed your practice, if at all?

Dr Agrawal:

So that's a great question and I think that is the crux of this research. And the bottom line that we want to get towards that is what is the clinical impact of this study and what we can take to our patients. And I think these data are rather convincing in favor of ileocecal resection as that being a good treatment option for the right patient. And by the right patient, I mean individuals who have limited ileocecal or terminal ileal Crohn's disease and patients who are motivated to undergo surgery, because of course this is surgery. And so in this right patient, I think that ileocecal resection is a viable option and it should be brought to the table to towards discussion with our patients. And we'd be surprised by how many patients are open to the idea of a one-time surgery and rather than having to undergo routine infusions and considering the risks of anti-TNF.

Of course this is not to say that following surgery patients are cured. Depending on the clinical features of a patient, we may determine that they warrant postoperative biologic therapy or they may warrant postoperative monitoring to look for endoscopic and clinical recurrence of disease. But even then, this is an important consideration and many patients may be in favor of it. This I think is especially important in settings which may be more resource-limited—for example, developing and recently developed countries where continuous or regular infusions may be cost prohibitive. And so in those scenarios where surgery may be more easily available, this could be a consideration. And having talked to colleagues from India, it seems like they're also very enthusiastic about ileocecal resection being an option considering that tuberculosis is a very important consideration in many of these countries, this can also be a factor against starting off anti-TNF right away. And so this gives our physicians and our patients more treatment options again in the right context.

Dr Cross:

Yeah, I think back, I'm much older than you, but when I was training there was this concept that if you sent your patient to surgery, it was a failure of the gastroenterologist. And I think at the very least, that's clearly not the case. So if you offer your patient a surgery and that restores their quality of life like this shows and like LIR!C shows, that is a completely reasonable option. I also think that we still sometimes because of that belief, maybe even unconscious, is that patients are cycling multiple therapies for deep-seated disease or complicated disease and that's clearly appropriate, inappropriate. So when I see a patient with stricturing disease or penetrating disease, if there's not stenotic dilation or not a big abscess fistula, I may offer them biologic therapy. And that could include of course an anti-TNF. But I make it very clear because I think the chances of long-term success in that situation is unlikely that I'm only going to try one biologic and we're not going to cycle through 2 or 3.

We'll give a chance with one and reassess and if everything's great, awesome, and if not, let's do a surgical reset. And I do think you can think of a lot of good examples. You mentioned resource-poor countries where this would be a good option. Someone who's about to go to college, maybe you catch them before they're going to go to college, this might give them that 5-year period where they're asymptomatic and don't have to worry about drug therapy. And the other message I got from ECCO is it seems like they do a lot less postoperative prophylaxis than we do in the US and I think probably we need to be somewhere in between. I think we probably overuse it a little bit here instead of doing aggressive monitoring. So I think this is really timely, it's really good and I look forward to more research on this to figure out maybe who the optimal patient is for surgery upfront.

One quick question, and I've grappled with this both of LIR!C and studies like yours is when you do a surgical reset, how do you adjust for that in time dependent analysis? It seems like you're almost biasing against the anti-TNF because there's going to be some disease there, even if they're not stricturing, they might be submucosal fibrosis present. So have you thought about that, how you would potentially do some kind of adjustment for that?

Dr Agrawal:

That's a great question. I think to account for that, the first distinction in my mind that I make is that surgery, following failure of medical treatment is very different from surgery as primary therapy in an individual who may be bionaive. And so even though the surgery is ileocecal resection, in both instances they're being done on very different patients and with very different disease activity and underlying complications and thereby different outcomes. And so I would almost treat patients in these two categories as having these two separate exposures as primary ileocecal resection or ileocecal resection after failure of medical therapy. I think that would help account for differences between the two to some degree. And then of course you have to account for the time interval between diagnosis and the ileocecal resection. So take into account as the variable as a time varying variable, and then ask the research question.

In our analysis, we restricted it because we wanted to get a very clean group, so we restricted it to only the very first surgery. And individuals who underwent subsequent surgeries were not included in the initial analysis. Subsequent surgeries were considered as outcomes and not as the exposure variable. But I completely agree with you that this is tricky. And this highlights the point that in outcomes research such as this, we have to be very thoughtful and very mindful of how we define all of our variables including exposure, outcome, variables, and covariates, because that would really influence how our results come out and how we can interpret them and put them in clinical context.

Dr Cross:

And you nicely highlighted that when you talked about potential limitations for these types of studies. So this is the best question. This is the fun question. So tell the listeners something about yourself that they may not know from looking at your bio profile. What do we need to know about Manc that's not in there?

Dr Agrawal:

I don't know if you need to know anything about me that's not in there, but a fun fact is that I like to run a lot and it's on my runs that I think about all the research that I'm doing and probably get my best research ideas.

Dr Cross:

Alright, well that's the way you focus. Great. Well Manasi, thanks for doing this. Congratulations on the publication and presentation at ECCO, and we hope to have you back on IBD Drive Time soon.

Dr Agrawal:

Thank you very much for having me.