Different Staging, Risk Stratification, and Treatment Methods for Acute GVHD

Case-Based Peer Perspectives Spotlight LiveNovember 2 CBPP Spotlight

During a Targeted Oncology Case Based Peer Perspectives event, Miguel-Angel Perales, MD, discussed testing for risk stratification and treatment of graft-versus-host-disease based on the case of a 48-year-old man.

During a Targeted Oncology Case Based Peer Perspectives event, Miguel-Angel Perales, MD, chief, Adult Bone Marrow Transplant Service, Memorial Sloan Kettering Cancer Center, discussed testing for risk stratification and treatment of graft-versus-host-disease based on the case of a 48-year-old man.

Targeted Oncology™: Can you describe the factors present that contribute to the development of GVHD?

PERALES: There are donor-recipient factors that we’ll see that have to do with HLA [human leukocyte antigen] disparity. In this case, the patient [and donor] matched. Sex matching is important in the case of female donors, [as are] donor parity and donor age. The ABO blood type is less of a factor, and the question of whether ABO [affects] survival or not has now been addressed in some large CIBMTR [Center for International Blood and Marrow Transplant Research], including a paper by Craig Kollman, PhD, from a few years ago.1 The main factors in that paper for survival were HLA and donor age; those were the 2 that stood out in the multivariate analysis.

The CMV serostatus is important, and there also [have been] data around cytokine gene polymorphisms, but this is not something that is part of routine clinical practice. When it comes to stem cell grafts, we know that PBSCs have a higher risk of GVHD than bone marrow. In most cases, bone marrow has a higher risk than cord blood; in many cases, the advantage of cord blood is that we can get around the HLA barriers. You can safely transplant patients with cord blood that may be [matched at a minimum of 4/6 loci]; this could even translate into a haploidentical graft in many cases without post-transplant cyclophosphamide and get comparable levels of acute GVHD.

In terms of the graft itself, higher CD34 counts have been associated with a higher risk as well as higher T-cell dose. Finally, the ablative conditioning regimen is also associated with more acute GVHD.

Would you perform a biopsy at this point to confirm GVHD?

[When asked what to do next], most [oncologists] would perform a biopsy to confirm GVHD and some people would test for biomarkers. Others would go straight to treatment.

The use of biopsies in GVHD is something that many of us strive for. The issue is that in many cases the biopsy comes back as nondiagnostic or the dreaded “could be GVHD” [results]. Most of us still try to get biopsies in some cases. It can help differentiate the patient’s condition from other potential etiologies. Particularly, you also want to rule out infectious etiologies, and some of those are best picked up with a biopsy.

What are the different staging systems for GVHD?

The MAGIC Consortium is familiar to most. The difference with [this system versus] the traditional grading system is that, in addition to the skin surface, bilirubin, and upper gastrointestinal symptoms, the MAGIC Consortium is able to estimate stool volume based on episodes of diarrhea. This is particularly useful in the outpatient setting where patients are not measuring their stool volume.

[In this patient’s case], you can use a rough estimate of what that corresponds to. We’re looking at a stage III [acute GVHD] in the skin. Calculating [based on] 200 cc per episode of diarrhea in this patient, there were 4 episodes of diarrhea per day over 2 days; that gives you a stage II in the lower gut.2

In addition to the traditional grading system and the MAGIC modifications, there have been a couple of other recent attempts to try to further define the severity of GVHD. The reason for that is we know that patients with different stages and grades of GVHD are going to respond differently. We’re particularly interested in understanding which patients are more likely to respond in those who have failed treatment. [In a study conducted by Margaret] MacMillan, MD, MSc, and colleagues, they defined 2 groups of acute GVHD: a standard-risk [population] and a high-risk patient population. It was a large data set of almost 2000 patients, and standard risk in this cohort represented 84% of patients. It was primarily patients with skin or early-stage gut involvement.3

From the same paper, by defining 2 different populations in the standard-risk group, the acute GVHD response to steroids at day 28 was almost 70% [95% CI, 66%-70%] with almost 50% [95% CI, 38%-50%; P≤ .001] complete remission [CR]. In contrast, in the patients who had high-risk disease—that was only 16% of the overall cohort—their chance of overall response was just over 40%, so 30% less [versus those with standard risk]. The CR rate was almost half of what you saw in the standard-risk group, so down to 27%. The risk of transplant-related mortality at 6 months went from 22% in the standard-risk group to 44% in the high-risk group.3

What is the importance of risk stratification in the treatment of patients with acute GVHD?

This type of differential risk stratification is helpful when you think about your newly diagnosed patient with acute GVHD, and particularly as we think about studies that risk stratify patients where we may want to go beyond the standard therapy with just steroids alone. One way to do that is to do a randomized study, steroids plus or minus drug X. Another approach would be to take patients who are high risk and say that, based on their outcome being much worse with standard steroids and their risk of non-relapse mortality being as high as 44% at 6 months, it’s reasonable to put those patients on a study asking questions about [the efficacy of] additional therapies in addition to steroids.

The other approach to risk stratification is the work of John Levine, MD; James L. Ferrara, MD; and Sophie Paczesny, MD, PhD, which they started in Michigan and more recently have continued at Mount Sinai in New York with the use of biomarkers. One of their early papers dating back to 2012 shows that they were able to stratify patients based on biomarkers. Ferrara has shown some elegant work taking certain biomarkers—and ST2 seems to be one of the most sensitive ones that you can take in patients with grade 2 GVHD—and clearly [separating] out patients who are going to behave [similarly] to those with grade 1 GVHD who have low ST2 levels, and patients who are going to do as [poorly] as patients with clinical grade 3 GVHD [with high] ST2 levels.4

Which agents would you consider using to treat here?

I usually reserve budesonide for upper gut GVHD, although you would think it was better for stage I lower gastrointestinal. In somebody who has 4 bowel movements a day, you’re already at stage II in the gut, so that’s probably somebody I would also treat with steroids. The dose you would use is 2 mg/kg. That’s the standard. There’s been a lot of work comparing 1 mg/kg versus 2 mg/kg, but I think many of us go for 2 mg/kg.

The goals of primary therapy for acute GVHD are to improve or stabilize organ manifestations of GVHD, limit some of the longterm treatment-related toxicities, and improve functional capacity or quality of life. First-line therapy remains treatment with steroids. Recently, we’ve seen the approval of ruxolitinib [Jakafi] as the first drug approved for second-line therapy.5 Ultimately, I think the goal for all of us is to try to improve overall survival; there is significant mortality [associated with acute GVHD].

How do you taper steroids once you achieve initial response?

I think anybody who has done trials of GVHD knows how hard it is to get consistent [drug] tapering across centers, even within centers [where] the physician practices. One of the biases that is often introduced—if you don’t give the patient a specific tapering calendar—is that the patients who are doing well are not seen as often and then are tapered less often than the patients who are doing less well and who are seen more often. Those patients may be tapered faster. There are a lot of biases introduced into the tapering schedule. In clinical trials, there’s been an effort to provide the patient with an up-front schedule of drug tapering and only deviate if they’re not responding.

What is considered a standard tapering approach is the [decrease of the steroid dose] 10% every 5 to 7 days as tolerated.

Early progression or failure to improve within 5 to 7 days, or incomplete response after 28 days, defines refractoriness or resistance to steroids. Steroid dependence is the inability to taper below 2 mg/kg or [the incidence of] recurrent flares during the taper. Then, intolerance is unacceptable toxicity.6

What are the guideline-recommended regimens for patients with steroid-refractory acute GVHD? What are the data behind them?

In terms of treatments, there are NCCN [National Comprehensive Cancer Network] guidelines. If you look them up, you’ll see that they’re not very detailed. They basically list a number of drugs that can be used. The only 2 drugs that are currently FDA approved for steroid-refractory GVHD [are] ruxolitinib for acute GVHD and ibrutinib [Imbruvica] for chronic GVHD.7

REACH-1 [NCT02953678] was a single-arm phase 2 study of ruxolitinib for steroid-refractory acute GVHD. [The results] were published earlier this year. Basically, patients developed acute GVHD and once they had been defined as being refractory or resistant to steroids, they could then be enrolled on the study. Ruxolitinib was started at 5 mg twice daily and could be escalated after 3 days to the full dose at 10 mg. The primary end point was overall response rate [ORR] at day 28. There were also a number of secondary end points.8

The ORR at day 28 was almost 55% [95% CI, 42.7%-66.8%] with a CR rate of 26.8%. The best overall response at any time during treatment was 73.2% [95% CI, 61.4%-83.1%]. The median time to response was 7 days. Patients responded quickly. The median duration of response was almost 12 months, and then relapse was seen in about half the patients, meaning recurrence of GVHD. The non-relapse mortality at 6 months was still quite high at 44.4% [95% CI, 32.5%-55.7%].9 These are all patients who failed steroids. On the basis of this study, the FDA approved ruxolitinib for steroid-refractory acute GVHD.2

Are there any randomized data supporting the use of ruxolitinib in these patients?

At the same time, the European Medicines Agency had requested a randomized trial, and [that resulted in the] phase 3 REACH-2 [NCT02913261] study design. [It had] similar inclusion criteria as REACH-1. Patients were randomized to either ruxolitinib at 0 mg twice daily versus best available therapy. The primary end point was identical, ORR at day 28, with similar secondary end points as the REACH-1 study.

This was a randomized phase 3 trial [that] led to approval in Europe for ruxolitinib in this indication and confirmed the United States data. The results are quite similar. The ORR at day 28 was 62% [versus 39% with the control (odds ratio, 2.64; 95% CI, 1.65- 4.22)]. The 27.9% CR [at day 28] rate was almost identical to what we saw in the other trial compared with control.10 The patients with ruxolitinib did better than the patients in the control arm.

If you’ve used ruxolitinib either in this setting or in other settings in patients with [hematological] malignancies, 1 of the main complications is myelosuppression. In the GVHD setting in particular, we see thrombocytopenia. Many of these patients start off with low platelet counts, but there is a slightly higher rate of thrombocytopenia with ruxolitinib. The other thing to be aware of is that there have been documented cases of CMV reactivation, and there appears to be a higher risk with ruxolitinib than with other therapies used in these patients.

The way these studies look at CMV infection is they lump together viremia and disease, so they did not separate the 2 out. We don’t have data on viremia versus disease. I think the other thing to take into account is that in the United States, we’ve now had letermovir [Prevymis] approved for [more than] 2 years. These studies—REACH-1 [and REACH-2]—were done in most patients [who did] not have access to letermovir prophylaxis. I think many of us think of letermovir as a good drug for CMV prophylaxis, particularly in patients who require steroids for GVHD. That’s something to consider that should decrease the risk of CMV.


1. Kollman C, Spellman SR, Zhang MJ, et al. The eff ect of donor characteristics on survival after unrelated donor transplantation for hematologic malignancy. Blood. 2016;127(2):260-267. doi:10.1182/blood-2015-08-663823

2. Harris AC, Young R, Devine S, et al. International, multicenter standardization of acute graft-versus-host disease clinical data collection: a report from the Mount Sinai Acute GVHD International Consortium. Biol Blood Marrow Transplant. 2016;22(1):4-10. doi:10.1016/j.bbmt.2015.09.001

3. MacMillan ML, Robin M, Harris AC, et al. A refi ned risk score for acute GVHD that predicts response to initial therapy, survival, and transplant-related mortality. Biol Blood Marrow Transplant. 2015; 21(4):761-767. doi:10.1016/j.bbmt.2015.01.001

4. Levine JE, Logan BR, Wu J, et al. Acute graft-versus-host disease biomarkers measured during therapy can predict treatment outcomes: a Blood and Marrow Transplant Clinical Trials Network study. Blood. 2012 April; 119(16): 3854-3860. doi:10.1182/blood-2012-01-403063

5. FDA approves ruxolitinib for acute graft-versus-host disease. FDA. May 24, 2019. Accessed October 20, 2020. https://bit.ly/2HgBNoQ

6. Schoemans HM, Lee SJ, Ferrara JL, et al. EBMT–NIH–CIBMTR Task Force position statement on standardized terminology & guidance for graft-versus-host disease assessment. Bone Marrow Transplant. 2018;53(11):1401-1415. doi:10.1038/s41409-018-0204-7

7. NCCN. Clinical Practice Guidelines in Oncology. Hematopoietic Cell Transplantation, version 2.2020. March 23, 2020. Accessed March 23, 2020. https://bit.ly/3ocIjNV

8. Jagasia M, Zeiser R, Arbushites M, et al. Ruxolitinib for the treatment of patients with steroid-refractory GVHD: an introduction to the REACH trials. Immunotherapy. 2018;10(5).391-402. doi:10.2217/imt-2017-0156

9. Jagasia M, Perales MA, Schroeder MA, et al. Ruxolitinib for the treatment of steroidrefractory acute GVHD (REACH1): a multicenter, open-label phase 2 trial. Blood. 2020;135(20):1739-1749. doi:10.1182/blood.2020004823

10. Zeiser R, von Bubnoff N, Butler J, et al. Ruxolitinib for glucocorticoidrefractory acute graft-versus-host disease. N Engl J Med. 2020;382(19):1800-1810. doi:10.1056/NEJMoa1917635

Recent Videos
Corey Cutler, MD, MPH, and Hannah Choe, MD, experts on GVHD
Corey Cutler, MD, MPH, and Hannah Choe, MD, experts on GVHD
Corey Cutler, MD, MPH, and Hannah Choe, MD, experts on GVHD
Hannah Choe, MD, an expert on GVHD
Corey Cutler, MD, MPH, an expert on GVHD
Corey Cutler, MD, MPH, and Hannah Choe, MD, experts on GVHD
Corey Cutler, MD, MPH, and Hannah Choe, MD, experts on GVHD
Guillermo Garcia-Manero, MD, an expert on MDS
Related Content