Yi-Bin Chen, MD: There are certain biomarkers that have been developed for acute graft-vs-host [GVHD] disease. None of them are diagnostic. The 1 biomarker that can be sent off right now that is commercially available is called the MAGIC [Mount Sinai Acute GVHD International Consortium] algorithm probability, or the MAP. That’s based on the measurement of 2 proteins, REG3A and ST2, which are basically signals of GI [gastrointestinal] mucosal injury. It is not diagnostic because any transplant patient who has significant complications early on will have elevated REG3A and ST2 levels, and thus an elevated MAP score. It is meant to be prognostic in the setting of a clinical diagnosis of acute graft-vs-host disease [that has] already been made.

There have been several large studies, mostly done by the MAGIC consortium, of which we’re members, that have studied the use of biomarkers in acute graft-vs-host disease. The utility exists, or at least has been shown to possibly exist, in several settings. Measurement at 1 week after transplant appears to be able to risk-stratify patients who are at high risk of developing severe acute graft-vs-host disease vs those who are not. That’s prior to the development of and gets into the insight into the biology or the immune cascade leading to acute GVHD.

Perhaps best studied is at the onset of acute graft-vs-host disease, when patients [first] present [with the disease]. That’s led to a certain platform of possibly being able to risk-stratify patients at the diagnosis of acute graft-vs-host disease, and thus [conduct] trials where you take high-risk patients and give them 1 treatment, and low-risk patients and give them another treatment. The biomarkers have also been validated at 1 week after starting treatment to help predict clinical response and tease that out. If your patient’s diarrhea hasn’t improved at 1 week, could we send the biomarkers to help us understand if they can better predict if they will or will not, because measurement of diarrhea seems a bit crude to try to figure this out?

Then at 1 month after, or day 28, after treatment has started, it’s been helpful to predict long-term outcomes. There have been no prospective trials that have said that the measurement of these biomarkers in any setting should dictate any decisions that we make. That has never been shown. We are actively conducting clinical trials to try to prove that, but these biomarkers exist to be able to be measured and so forth. I fear that some are sending them as a diagnostic biomarker, though I don’t know that, and that would be inappropriate use of such. I do support the use of these biomarkers in clinical trials to figure out if they will indeed be truly useful.

I should mention there’s 1 biomarker that was presented at recent meetings by Shernan Holtan, [MD,] and her colleagues at the University of Minnesota called amphiregulin. That is a ligand for the EGF [epidermal growth factor] receptor. They’ve shown that this biomarker [when] measured at the beginning and 8 weeks later at the start of treatment seems to correlate with outcomes in acute graft-vs-host disease. This is being further validated and compared to the MAGIC biomarkers as well to see if incorporation of all 3 will improve, or if it doesn’t add any utility at all. We look forward to all of these investigations to see how we can use these biomarkers.

Corey Cutler, MD, MPH, FRCPC: Linda, do you and your colleagues use these biomarkers routinely for the prognostication or the treatment of your patients with acute graft-vs-host disease?

Linda Perry, PA-C: We do, mostly with relation to clinical trials. Hopefully, we’ll be able to use them more on a regular basis outside of the trial setting, but mainly as part of clinical trials and such.

Corey Cutler, MD, MPH, FRCPC: It’s the same thing at our center. We’re not sending them on patients who are not participating in clinical trials at the moment because as Yi-Bin said, we’re not sure what to do with them, but certainly, we’re using them for risk stratification.

Transcript edited for clarity.