Steroid-Refractory Chronic GVHD: Management Strategies

Yi-Bin Chen, MD: We’re going to transition to talking a little about those newly approved agents. We’ve had 3 approvals in the last few years for steroid-refractory chronic graft-vs-host disease [GVHD]: ruxolitinib [Jakafi], ibrutinib [Imbruvica], and belumosudil [Rezurock]. We’re going to talk a little about each one. The most recent approval was ruxolitinib. You heard about it being approved for steroid-refractory acute, and most recently, it was approved in this past year for steroid-refractory chronic graft-vs-host disease. Hannah, do you want to talk a little about the pivotal data that got ruxolitinib approved?

Hannah Choe, MD: Sure. Ruxolitinib for chronic graft-vs-host disease was studied in the corticosteroid refractory setting in a phase 3 trial with patients with moderate to severe chronic graft-vs-host disease. They hadn’t received a heavy pretreatment burden, so most of the patients had only received corticosteroids or calcineurin inhibition. Only roughly 14% to 15% of them had received an additional systemic therapy on top of that. Of the adult population studied, I believe in patients 18 years and older, the patients were randomized to ruxolitinub vs best available therapy. Best available therapy in this case was an investigator’s choice of 10 commonly used immunosuppressive or chronic GVHD therapies, which for the most part was ECP [extracorporeal photopheresis], mycophenolate, and ibrutinib, but also included infliximab [Remicade], pentostatin [Nipent], etc.

What’s great about that trial is that it was a pivotal trial because it had such an extraordinary overall response rate in corticosteroid refractoriness of roughly 50%. They also looked at failure-free survival and saw a major difference for patients, 18.6 months vs 5.7 months, in best available therapy. There was a very evident improvement for patients in all organ systems as well. It was the first big step for chronic graft-vs-host disease to finally get phase 3 level support for FDA [Food and Drug Administration] approval.

Yi-Bin Chen, MD: Yes. All of us were impressed by the large phase 3 trial, the coordination, and the investment doing so.

Sophie Paczesny, MD, PhD: And they extended the age range to 12 years. It’s 12 years and up for approval. That’s pretty impressive.

Hannah Choe, MD: Yes, you’re absolutely right. I’m sorry.

Yi-Bin Chen, MD: Yes, it’s pretty impressive. For us, it’s become our standard for second-line care for chronic graft-vs-host disease. But several years prior, we had the first approval, which was in the oral agent ibrutinib. Sophie, do you want to talk about what the mechanism of action is and what the results have been for ibrutinib?

Sophie Paczesny, MD, PhD: Ibrutinib inhibits both ITK [inducible T-cell kinase] and BTK [Bruton tyrosine kinase]. Basically, ITK is on the T cell and BTK is on the B cell. That’s the mechanism of action. It was the first FDA approval based on a small clinical trial, with some efficacy in the second line. That was the first improvement in the first FDA approval drug in the field, even before acute GVHD. It’s maybe a little more difficult to give, and there’s some toxicity, so it seems that ruxolitinib is better tolerated and people will go more often to ruxolitinib rather than ibrutinib. In pediatrics, ibrutinib is still used and can help some patients.

Yi-Bin Chen, MD: Yes, you brought out a very good point, that the pivotal trial only had 42 patients and it was a single-arm trial with some impressive efficacy, but when looking back, there were some more stringent eligibility criteria. There have been a couple of real-world analyses published using ibrutinib that have illustrated what you’re saying, that it’s a bit more toxic and immunosuppressive than other agents. We tend to use more ruxolitinib than ibrutinib. The third agent, belumosidil, was approved in 2021, too. It has a novel mechanism of action. Nelson, do you want to talk about that a little?

Nelson J. Chao, MD, MBA: Belumosidil is an oral medication that’s a ROCK2 inhibitor. ROCK2 is a Rho-associated coiled-coil containing protein kinase 2. ROCK2 is a profibrotic stimulus. It stems profibrotic activity. The inhibition of ROCK leads to a drop in STAT3 [signal transducer and activator of transcription 3] and a balance to STAT5 [signal transducer and activator of transcription 5]. As Sophie was saying earlier, that has an implication on the T cell side of things, where it inhibits Th17 [T helper 17] cells, which are pro-inflammatory cells, and probably helps shift the balance from Th17 cells to Tregs [regulatory T cells]. This goes back to the theme that you’re inhibiting these Th17 pro-inflammatory cells at the same time as it has an impact on fibrosis. Fibrosis is one of the most common hallmarks of chronic graft-vs-host disease.

The study was well designed, but it wasn’t a randomized trial. The investigators didn’t feel it could be randomized because these patients had failed at least 2 different lines of therapy. The data were pretty encouraging; more than two-thirds of the patients responded. What was most interesting to me was that the toxicity profile was quite impressive. It was pretty well tolerated. As with ibrutinib, we have to see what it’s going to look like in the real world as people start to use it.

Yi-Bin Chen, MD: Yes. There’s certainly more real-world experience with ruxolitinib than with belumosudil just because it was available even before approval. It’s interesting; if you look at all these trials, they list each of their patient characteristics, and all of them have different organs in involvement, and all of these agents show responses in all the different organs. All of these agents in their trials—at least for belumosudil and ruxolitinib—showed responses in patients who had failed ibrutinib in the ruxolitinib trial, as well as ruxolitinib in the belumosudil trial. That says to me that patients will respond, but we don’t know how to predict which patient will respond to what. Sophie, is there a way that we’re going to make progress here?

Sophie Paczesny, MD, PhD: First of all, second line is probably too late. When the NIH [National Institutes of Health] consensus say that it would be interesting to do an adaptive trial and try early on adaptive trial, you can try a different drug early and change and see what’s working best. You can also use biomarkers to help you; that will be key, too. We shouldn’t wait too long. Maybe at diagnosis, we could do something similar to the BMT CTN 1501 trial. This hasn’t been done yet with biomarkers and clinical characteristics. We can also maybe try something just before the second line, when beyond refractory, trying to see also biomarkers and compare these patients in an adaptive trial. That’s what I suggest, which is basically what the NIH consensus recently suggested.

Nelson J. Chao, MD, MBA: The data are good. These are the types of things that machine learning could help if you have 30,000 or 40,000 patients with reasonable, good data.

Sophie Paczesny, MD, PhD: That’s interesting. You need the patient numbers.

Nelson J. Chao, MD, MBA: But the data is there. We just haven’t learned how to look for it.

Yi-Bin Chen, MD: It’s a question of whether you think clinical characteristics will predict, or if we need more samples and biological characteristics. I’m not sure. Hopefully we’re able to do machine learning with models. Hopefully we’re able to get the samples and try to figure this out to be more precision medicine in some way with this treatment as well.

Transcript Edited for Clarity