Implications of the GEMINI Trial in IBD

Transcript:

Miguel Regueiro, MD: Bill, I'm going to pivot a little bit and move to vedolizumab and some of the work you did in post hoc analysis with the GEMINI-1 study, but now looking at deep remission with vedolizumab for patients with moderate-to-severe ulcerative colitis [UC]. First, tell us a little about the study design outcomes and then what this means to you in your practice.

William J. Sandborn, MD: It's interesting, the things that Marla and Jessica were just saying. Our group [at the University of California San Diego], it’s on the website, had an abstract for DDW [Digestive Disease Week] where we looked back. I had been practicing treat-to-target since I came to the University of California San Diego almost 10 years ago. We had all these patients for whom we were doing treat-to-target for endoscopy; I was systematically biopsying them along the way, but not treating to biopsy. A subset of the patients who got endoscopic healing got to histologic healing and a subset didn't. You could look at the outcomes of hospitalization, surgery, and complications. Getting to histologic healing, incidentally, when you were aiming for endoscopic healing, the patients did significantly better than if you didn't get to histologic healing. From an association standpoint, it's better. As you're selecting drugs, if you know what the combined histologic and endoscopic healing rates are, it may give you greater confidence that you're going to get to where you want it to go.

Marla said a couple of interesting things. One is that a good bit of work needs to be done to understand if the hypothesis that she just raised is right: that fecal calprotectin is a stool histology because we've known that fecal calprotectin in ulcerative colitis only correlates with endoscopy about 80% and is probably 70% or less with Crohn disease, and we thought it was a false negative or positive. Maybe it's not; maybe it's an accurate predictor of histology, which also doesn't perfectly correlate with endoscopy. That'd be really cool.

Then, the other thing is the treat-to-target and the cost effectiveness. We've been looking at that. As you go through the fourth or fifth ratchet, the incremental benefit is pretty low because the drugs that we have as a group across mechanisms are not that effective when you get into multiple failures. You get all this extra cost in your dose escalating, and you're not getting a lot of incremental benefit. Getting to histologic healing gives you better outcomes, but you've got to trade that off against some level of social responsibility for not prescribing expensive stuff that's unlikely to get you where you want to go.

That's an interesting set of ideas. Coming back to your question about vedolizumab, what we found was that, in ulcerative colitis, you get a deep remission as defined by endoscopy and clinical symptoms together such that the patients clearly do better. They have higher long-term, steroid-free clinical remission rates, fewer complications, less hospitalization, and better quality of life. Patients who get to deep remission do better, which is what we would have intuitively expected, but it was gratifying to see that.

David T. Rubin, MD: Miguel, may I make an additional comment? I may join the discussion here. I wanted to add that there are many things that we still don't understand about histology, not the least of which is how it changes over time. When we talk about a cross-sectional moment of assessment, we don't actually appreciate how much of that may be going up and down as the patient's disease changes and how much of it may lead relapses, even though it seems to make good sense. I endorse everything my colleagues have said. I also happen to think that when calprotectin is elevated, it's a predictor of relapse if you de-escalate or de-intensify treatments. We've seen that in both Crohn disease and UC, but I acknowledge that there's a lot we don't know, and that is what Bill mentions in terms of it not always aligning with endoscopy, nor does it always align with symptoms. Nothing is perfect, of course, but it's important to keep those limitations in mind.

I'll add that we can keep looking deeper. There's a study we [at the University of Chicago Medicine]

Did years ago now with PET [positron emission tomography] scans in ulcerative colitis, and all the people who we did PET scans on had an endoscopy, biopsies, and clinical assessment that said they were in deep remission and healed, yet 4 out of 11 patients had PET scans that lit up in their sigmoid colon despite that, suggesting functional activity of some kind or metabolic activity. There are a lot of different ways to continue thinking about this, and we don't want perfect to be the enemy of good enough. We have to keep thinking carefully about this.

William J. Sandborn, MD: David, there's some dumb stuff that is not well standardized, so how many biopsies do you take, and where do you take them?

David T. Rubin, MD: Right.

William J. Sandborn, MD: Our group has been looking at that recently. If you compare 1, 2, 3, or 4 where you plateau, and it looks like within a segment, you plateau at 2, and 2 is about is good as 3 or 4. But then if you look across segments, there's some variation. Do you look at the average of multiple segments? Do you look at the worst segment? Do you pick a segment, like the sigmoid colon and follow that in all the patients, or the rectum if that's the proximal extent of the disease? Lack of standardization in all those things, both in clinical practice and to some degree in clinical trials, adds an element of heterogeneity to all that we're looking at, and we desperately need some standardization.

David T. Rubin, MD: I agree completely. I'll get into the weeds with one last comment before I let Miguel take over, which is that when you use the microtome as you go 3-dimensionally through fragments of biopsies, you get different levels of histological activity, so it's gets even more complicated.

Miguel Regueiro, MD: It’s going to be a long way to go in our biomarker world, which includes histology, and we still need the perfect biomarker, especially in this time of COVID-19 [coronavirus disease 2019], when we're not bringing patients in and trying to figure out what their fecal calprotectin is and bloodwork, even if they get that.

Transcript Edited for Clarity