Current Practices and Future Directions for Treatment of Graft-vs-Host Disease (GVHD) - Episode 7
Richard Maziarz, MD, provides an overview of the drug ruxolitinib, which was recently approved for use in steroid-refractory acute GVHD (graft-vs-host disease). He leads a discussion on the results of REACH-1 and REACH-2 and the experiences of the other panelists in using this drug.
Yi-Bin Chen, MD: Why don’t we talk about acute GVHD [graft-vs-host disease] and what’s happened with the therapy for steroid-refractory acute graft-vs-host disease?
Richard Maziarz, MD: We’re excited to see the data from REACH-1 and REACH-2; I know you were a participant in the REACH-1 study at Mass General [Massachusetts General Hospital]. This utilized the drug ruxolitinib; it now has an FDA label for steroid-refractory acute GVHD. That is a major milestone. One of the aspects is that a single agent is going to affect multiple biologic pathways; in that way, we should theoretically enhance our ability to treat GVHD. When you examine some of the pathways on how ruxolitinib could benefit a patient with steroid-refractory acute GVHD, it’s been shown that it will decrease the expression of the signaling molecules on antigen-presenting cells. It blocks MHC class II expression, and the JAK-STAT [Janus kinase-signal transducer and activator of transcription] pathway is the intracellular common pathway for multiple extracellular growth factor cytokine cell receptors; you can simultaneously block multiple cytokine stimulation. It’s been shown to block effector T-cell expansion and enhance regulatory T cell; it has multiple pathways from a single drug that could theoretically change the course of treatment.
The multicenter phase 2 study demonstrated that, in a setting where we would have anticipated a lower response rate historically, we now saw approximately 55% overall response [rate], which was higher than we expected. It was a day-28 response, but then at 6 months, it was sustained.
Corey Cutler, MD, PhD, FRCPC: I agree, it is a major milestone. However, the devil is in the details here. The CR [complete response] rate in the randomized REACH-2 trial was only 25% in patients with grade 3/4 disease. At the end of the day, there was no survival advantage in the ruxolitinib arm. From our point of view, while it is great to have a drug to say you have failed steroids, this is the next agent and this has some demonstrated activity, we are not done yet with a 25% CR rate in grade 3/4 disease. By 1 month, a third of the patients who had a response had lost it by day 56. Therefore, it is a tremendous step forward, but we need to see all of these other programs through to their natural conclusion to make sure we can find other things that work in this setting.
Richard Maziarz, MD: I would concur, there is no change in overall survival. There was no enhanced relapsed rate, so people were still dying of nonrelapse mortality, but they were living longer; they did survive longer to get their opportunistic infection. It leaves the opportunity to create that steroid-sparing next step or intervention.
A paper that was published about 6 months ago showed that this is now the only drug that’s FDA approved for steroid-refractory GVHD. The study that was published said we now need to establish a ruxolitinib-resistant steroid-refractory GVHD—how we’re going to diagnose it and how we’re going to categorize it; that will be our next set of interventions.
Yi-Bin Chen, MD: I agree with both of you. It’s great to see the cooperative large trials done and to see the results, especially in the randomized setting as well, with the control being dealer’s choice. However, I’m not sure there’s a better control than that. Clearly, this is not a home run. It is probably where it should be, unless you guys disagree. It’s at least the standard of care now for steroid-refractory acute GVHD. All of us probably have other things we have always used for, say, bad GI [gastrointestinal] cases, that we believe in, but we tend to add it early, and even in combination for first sign of steroid refractoriness. Would you guys agree?
Richard Maziarz, MD: I would agree. It’s interesting, from the REACH-2 study, the randomized trial against best available therapy, that the population that gained the greatest benefit and had the greatest response rate was the most advanced patient. For a patient with grade 4 GVHD of the gut, if you could get the drug in there early, it could potentially assist in shutting down. I suppose if we looked at these steroid-dependent, steroid-refractory patients, or the patient who is at 7 days and we’re not making progress, there’s less to be gained there. It was those highest risk patient populations that we seem to dent, and the heterogeneity of the steroid-refractory population also dims a potential success that can be achieved with the drug.
Sophie Paczesny, MD, PhD: I did like the idea from Corey: to not wait too long and then to …. over drugs. Maybe this is where you want to have the cellular products on board, like MACs or regulatory T cells and things like that.
Yi-Bin Chen, MD: The success of ruxolitinib moves the field forward in the sense that we now have an accepted control arm for future studies. We have a platform that seems safe to combine with other agents, though that remains to be seen. Toxicity-wise, it seems like we could do the combinatorial strategies that Corey mentioned to think about moving the field forward. For the third line and fourth line steroid-refractory cases, I’m not sure a therapy exists that can turn the majority of those cases around. There is a loss of intestinal stem cells and localized thrombotic microangiopathy, and probably some organ damage that we can’t reverse. At that point, it is no longer an alloimmune response we’re trying to stop, but organ-specific inflammation that we can’t get control of. The key to acute GVHD, outside of prevention, is to control it early. I am not sure if second-line therapy is too late, but we can start to think about some creative approaches to take care of and treat patients before we get to that refractory point.
Transcript edited for clarity.