Yi-Bin Chen, MD, discusses the nuances of diagnosing patients with cGVHD, the need for biomarkers to inform who is likely to develop the disease, the introduction of ruxolitinib and belumosudil to the paradigm, and future directions with novel strategies in the space.
The approvals of ruxolitinib (Jakafi) and belumosudil (Rezurock) have shaken up the second- and third-line treatment of patients with chronic graft-vs-host disease (cGVHD), said Yi-Bin Chen, MD, who added that real-world data with these agents could further inform their utility in clinical practice and whether there is a place for frontline use in combination or in place of systemic steroids.
“Enough time has passed where we will see more real-world data for ruxolitinib in cGVHD to get a better handle on what is actually happening,” said Chen, director of the BMT Program and the Cara J. Rogers Endowed Scholar at Massachusetts General Hospital, and an associate professor of medicine at Harvard Medical School. “We all trust randomized trials, such as REACH3 [NCT03112603], but we do appreciate real-world experiences as we apply it because the majority of our patients don’t fit the eligibility [criteria] for how these trials are designed.”
“It’s too early to get any data on the real-world experience of belumosudil, but in the next couple of years that should be coming,” Chen added.
In an interview with OncLive®, Chen discussed the nuances of diagnosing patients with cGVHD, the need for biomarkers to inform who is likely to develop the disease, the introduction of ruxolitinib and belumosudil to the paradigm, and future directions with novel strategies in the space.
Chen: The diagnosis of GVHD, be it acute or chronic, is still a clinical diagnosis that requires the [patient’s] entire history, their symptoms, exam findings, and laboratory diagnostics. Although we can biopsy affected tissue—and those histological findings can help support a diagnosis of GVHD or rule out other things—the diagnosis remains the entire clinical picture.
There are times when it’s a bit difficult to know what is causing something for a patient. Of anything that happens to our patients in at least the first 2 years after transplant, cGVHD is always in the differential; we have to think about it. The disease is unbelievably heterogenous from patient to patient, so we can’t read a textbook to understand what we are going to see.
The manifestations are protean; it can affect almost any organ in the body. It’s interesting because cGVHD is not only a disease of the immune system and direct attack of that immune system on the patient, but it reflects a state of immunological disarray and probably causes that as well. [Patients] lose their normal regulatory mechanisms.
Therefore, we see some manifestations that are classical cGVHD where the donor immune system attacks recipients, but there are other manifestations that are just a reflection of immunological disarray. Those manifestations aren’t necessarily GVHD in the classical sense but are things such as immune-mediated thrombocytopenia or nephrotic syndrome. Those aren’t thought of as classical manifestations of the donor attacking the host, but they are evidence that the immune system has gone awry.
We pay attention to everyone who gets an allogeneic transplant. However, we must think about certain patients who are more at risk than others. That’s based on certain clinical risk factors that are inherent to the patient, the disease, the transplant platform used, GVHD prevention, and the events that happen afterward.
To take that in order, we know that certain characteristics of a patient make them more at risk for cGVHD. Specifically, the older in age someone is, if they are male and receive a graft from a female donor, and the actual transplant platform [are all risks]. If we use myeloablative conditioning vs reduced-intensity or nonmyeloablative therapy, which is certainly much more borne out for acute GVHD [aGVHD], it is a signal for cGVHD, especially [regarding] the use of high-dose total body irradiation.
Certainly, the type of donor matters in terms of donor and host [human leukocyte antigen] matching. That by far predicts for the most GVHD relationship. The graft source [also matters]. We know from randomized studies that peripheral blood stem cells have a much higher incidence of cGVHD vs bone marrow or cord blood if we can get [patients] past the early complications.
GVHD prevention [is critical]. We know that the inclusion of polyclonal anti–T-cell globulins, such as ATG products reduce the incidence of cGVHD. Ex vivo manipulations, such as T-cell depletion, do that as well. If we use the newly emergent, posttransplant, high-dose cyclophosphamide-based GVHD prevention [strategy], that appears to reduce the incidence of cGVHD as well.
By far, if we play all this out, the biggest risk factor along the way would be the development of aGVHD. That’s just a biological reflection of allogeneic reactivity. That is how we think of patients and what we have done. For a patient with a benign disease, such as aplastic anemia, I would design a transplant platform that would maximize GVHD prevention. I would know that patient is at a much lower risk [of developing GVHD] compared with someone with a high-grade hematologic malignancy who has a higher rate of relapse.
[In that situation], I would at least tolerate some GVHD to bring about better graft-vs-leukemia. That is how we think about patients as we go forward. We make conscious decisions about how we manage their immunosuppression in terms of rapidity of taper [of steroids] and things like that.
Going forward, it behooves us to invest in studies to try to define better ways to [make those decisions], not just with the diagnosis, but in real time along the way. That would [mean identifying] cGVHD biomarkers. That could help risk-stratify patients in real time to high or low risk for developing cGVHD. That would be a huge step forward and ongoing research is already trying to figure this out. It’s much more along for aGVHD vs cGVHD for a lot of reasons, but [ultimately] the key is to have biomarkers that can predict for certain manifestations. We could then either design preemptive approaches to prevent [cGVHD] or predict for differential responses to therapies. That way we could tailor our therapies based on those biomarkers.
In patients who present with localized cGVHD, which is very common, we tend to try to maximize local therapy. That includes eye drops to the eyes, topical mouth rinses for mouth involvement, and topical skin cream for limited skin involvement. At some point, if the manifestation becomes so severe or there is clear multi-organ involvement, [treatment] becomes much more of a systemic process than initially appreciated. There comes a time when we must decide to initiate systemic therapy.
Unfortunately, systemic therapy at this point remains the reinitiation of systemic steroids, such as prednisone. Most of these patients are outpatients [and receive] prednisone in oral tablets. We tend to give [prednisone] in a dose somewhere between 0.5 mg/kg and 1 mg/kg of the recipient’s body weight per day. That has been the same for the past couple of decades.
It is heartbreaking to add steroids back to a patient’s treatment because many patients have already been through a lot of complications like aGVHD or infections. Certainly, we need to treat the cGVHD, but we also know the patient is accruing some morbidity and risk [with additional steroids].
I worry about infections, causing or worsening diabetes, bone health and osteoporosis, worsening hypertension, and mood effects from long-term prednisone. [We also worry about] the effects on body habitus in terms of fat distribution and cushingoid appearance. All of those [risks] make steroids not an ideal choice.
We don’t have any data to say whether we should use a different agent for first-line therapy or combine steroids with something else; however, those trials are ongoing, and we are awaiting the results. More than half of patients will have an unsatisfying response to steroids, be it no response or a plateaued response that is not good enough for their quality of life. In more than half of patients, we need to add something [to steroids].
It’s sobering that recent data from the cGVHD Consortium would suggest that in anybody who we start on systemic steroids for cGVHD, only about a third of them ever get off steroids for a durable time. That is a reality that makes us think about how to frame [treatment] for patients in terms of long-term expectations.
In terms of second-line therapy, we view GVHD in general as an immunological disease, which it is. The historical therapies were immunosuppressive to treat the GVHD. Although that works to a certain extent, it comes with a high cost––that being certain opportunistic infections.
However, just as in aGVHD, the evolution in cGVHD [treatment] is toward more pathway-specific inhibitors that we think are especially active. Those [agents] also have less off-target toxicities, thus they yield less risk and morbidity to our patients who are a bit fragile.
We have some more options, excitingly. A few years back the BTK inhibitor ibrutinib [Imbruvica] was approved in steroid-refractory cGVHD. That [approval] was built on the foundation that there is some B-cell activity in cGVHD. For a subset of patients, that is why ibrutinib does work to a certain extent.
Other agents, such as rituximab [Rituxan], have had some efficacy. However, the real-world experience with ibrutinib has not mirrored what we saw in the clinical trial that led to its approval. That trial was somewhat restrictive in its eligibility criteria, so [the lack of consistency with the real-world experience] makes sense. That real-world experience has been a bit difficult not only in efficacy but just in the inherent toxicities [associated with ibrutinib].
More recently, ruxolitinib was approved for second-line therapy. This was based on the findings from the REACH3 study, which was one of the only large, randomized phase 3 studies in steroid-refractory cGVHD. [The results] showed a significant benefit with ruxolitinib vs best available therapy. The primary end points were overall response rate at 24 weeks and failure-free survival. Ruxolitinib had a significant benefit [in those end points] compared with best available therapy.
We’ve been using [ruxolitinib] at Massachusetts General Hospital as our standard second-line therapy for the past 4 years for cGVHD. This was partly because of the anecdotes that had been published, partly from our own experience, and partly because we have a lot of experience using ruxolitinib in aGVHD and other indications in which it is approved for. We had found it to be a very satisfying drug, although it is not a “home run.” We don’t have a home run if we look at all the trials to date for cGVHD. Of responses reported, complete remissions are the extreme minority. That’s how we know we don’t have a home run.
There is more work to be done in diagnosing cGVHD earlier and designing better therapies. Anecdotally, we have found that ruxolitinib benefits at least three-quarters of patients started on it. It is a well-tolerated agent, but we must watch for cytopenias. If a patient develops [adverse effects (AEs)], they can be managed with dose adjustments or dose interruptions. I hope one day to be able to see how [ruxolitinib] functions as first-line therapy or even in prevention [of cGVHD].
[In July 2021], belumosudil, an oral ROCK-2 inhibitor, was approved for third-line therapy of cGVHD. That [approval] was based on the momentum of the ROCKstar study [NCT03640481], which showed a response rate of over 70% with 2 different dosing strategies. Belumosudil is interesting in terms of its mechanism of action, in that it targets the fibrosis cascade. It certainly appears to help most patients without a significant amount of concern over toxicity. I want to see more data on the real-world experience with belumosudil in patients.
A couple of agents are being studied. We shouldn’t forget about extracorporeal photopheresis [ECP]. That is a modality that has been around for a long time and certainly has efficacy in cGVHD without much in the way of immunosuppression. The reason why it is not as attractive as the other agents is the logistics it takes for patients to receive the therapy in terms of proximity, time spent at the center, and [the need to insert] an indwelling catheter. Learning how to sequence ECP with [systemic] therapies is a must going forward.
A monoclonal antibody against CSF1R is being developed by a company called Syndax. That agent is being [tested] as an every-other-week infusion and targets activating macrophages that are thought to be very active in the fibrotic process, which is the histological hallmark of GVHD. Large expansion trials are ongoing right now [with that agent, called axatilimab]. We are all eagerly awaiting the results.
Outside of that, other agents are very early in development and are not yet ready for prime time. We saw an interesting abstract from Europe using a form of arsenic as frontline therapy that showed a very high response rate when combined with steroids. That goes along with what the next step is in cGVHD, which is to move approved therapies up to study them with systemic steroids in the frontline setting. This could not only improve response rates but allow more rapid steroid tapers to spare patients [from AEs] during response.
We also want to think about designing trials where we can parse out a subset of patients that don’t need steroids because they are low risk or [we] don’t think that that phenotype or biological sign of patients will respond to steroids. [In these populations], we can try these newer agents as frontline therapy without steroids. A goal is to replace steroids in the frontline setting, so these are trial designs being talked about in the community going forward. Now, one big step is already out of the way because these novel agents are approved in later-line settings.
I’m always interested in the newer agents being studied. These are small trials out of single centers, but those start the whole process. I am keeping my eye out for those types of studies.
As we move forward, it will be less about designing more therapies for third-, fourth-, or fifth-line treatment and more about how we can better prevent cGVHD either by doing so preemptively or by adding to the prevention backbone without sacrificing other factors. That has been much of the discussion in the community to try to glean whether there are any data we are going to see that will inform us to design those steps going forward.
Now that ruxolitinib is approved and a lot of us are using it as our standard second-line agent, we are going to see a lot of different formulations of ruxolitinib or other JAK inhibitors. In the community, we have been asking for about 3 years for some form of JAK inhibitor eye drops, skin creams, and mouth rinses. Right now, we use steroids for those topical formulations. Not only do we think that some sort of localized ruxolitinib application would be more effective, but certainly less toxic [vs steroids]. We will be watching out for that as adjunct treatment for the populations that have limited disease and don’t need systemic therapy.