Biomarkers Stratify Patients With Acute Graft-Versus-Host Disease

By Will Boggs MD

January 05, 2015

NEW YORK (Reuters Health) - A score based on three biomarkers can classify patients with acute graft-versus-host disease (GVHD) into clinically meaningful risk strata, according to research from the Blood and Marrow Transplant (BMT) Clinical Trials Network.

"BMT physicians have long desired a lab test that would distinguish patients with acute GVHD who would respond to steroids from those would not," Dr. James L.M. Ferrara from the University of Michigan in Ann Arbor told Reuters Health by email. "This algorithm does that for about half of all patients with acute GVHD."

Only a third of patients with acute GVHD achieve durable responses to initial corticosteroid therapy, the primary treatment of acute GVHD, and survival of those who don't respond is poor.

Dr. Ferrara and colleagues used the concentrations of TNFR1, ST2, and Reg3alpha, three recently validated biomarkers, to create an algorithm (the Ann Arbor GVHD score) that computed the probability of non-relapse mortality six months after GVHD onset.

Based on the algorithm, they assigned scores of Ann Arbor 1 to patients with predicted probability of non-relapse mortality of 10% of less, Ann Arbor 2 to patients with predicted probability of non-relapse mortality between 10 and 40%, and Ann Arbor 3 to patients with probability of non-relapse mortality of 40% or more.

When these scores were calculated for 792 patients, the responses to therapy differed significantly for score 1 (81%), score 2 (68%), and score 3 (46%)(p<0.0001 for all comparisons), according to the December 23 Lancet Haematology online report.

Durable responses were also significantly less likely in patients with Ann Arbor 3 GVHD than in patients with Ann Arbor 1 GVHD, regardless of organ involvement at GVHD onset.

Among patients who presented with skin GVHD and subsequently developed lower gastrointestinal GVHD, Ann Arbor score predicted the development of gastrointestinal GVHD, whereas the extent of skin rash did not.

Compared with Ann Arbor 2, Ann Arbor 1 predicted a lower risk of non-relapse mortality and Ann Arbor 3 predicted a higher risk of non-relapse mortality regardless of the presence of other clinical risk factors.

"This test has been validated for patients at the beginning of acute GVHD," Dr. Ferrara said. "We do not yet know whether serial algorithms of the same patient will provide useful guidance as to whether to intensify or taper therapy. That research is currently ongoing, but as of yet we have no evidence that the algorithm can be used that way."

"Additional ongoing research will also determine whether the algorithm can be used for patients who are at risk for GVHD but in whom there are not yet symptoms of disease," Dr. Ferrara said.

For now, Dr. Ferrara advised, "a patient with new acute GVHD symptoms (e.g. diarrhea) should have the test performed before the start of treatment. A patient with a low score (1) is likely to respond to standard therapy (systemic steroids), so steroids could be started, then tapered rapidly (over weeks rather than months) when the symptoms resolve. A patient with a high score (3) is not likely to respond to steroids alone, so that patient could immediately receive another therapy (e.g. TNF blockade, or anti-thymocyte globulin) in addition to steroids from the onset, preferably as part of a clinical trial."

"Patients with a middle score (2) have a fifty-fifty chance of responding to steroids alone, so the physician must rely on clinical judgment in those cases," he said. "The beauty of the algorithm is that these scores apply to all patients, regardless of the severity of their GVHD symptoms."

Dr. Ferrara added, "These markers are not (yet) routinely measured in bone marrow transplant (BMT) patients. They are currently available as a 'send out' to commercial labs like Viracor. But the ELISAs (enzyme linked immune assays) are very standard and very reproducible, and I expect that many hospitals with BMT programs will start to offer these tests soon."

"I've already had two emails from patients asking whether their doctors can send us the blood to do the test, so I guess there is indeed interest out there," he said.

In an editorial, Dr. Brian Betts and colleagues from Moffitt Cancer Center and University of South Florida in Tampa write, "This finding supports rational selection of high risk patients as defined by biomarker panel for novel targeted interventions."

They add, "We surmise that response to TNF inhibitors including etanercept or infliximab could be optimized among those patients with high serum concentrations of TNFR1."

"Overall, GVHD biomarker discovery is rapidly evolving," they conclude. "As a field, we must focus on functional and druggable targets, and ensure that the correct patient population is selected for clinical trials based on relevant biomarker expression."

Dr. Sophie Paczesny from Indiana University School of Medicine in Indianapolis recently reviewed GVHD biomarkers for the International Journal of Hematology. She told Reuters Health by email, "Addition of blood biomarkers to the classical GVHD severity score allows for better risk stratification at the onset of the GVHD clinical signs. The contribution of these blood biomarkers is expected to be significant, because the identification of steroid-refractory GVHD biomarker panels at symptom onset has tremendous potential for impacting our ability to risk-stratify patients before initiating GVHD treatment. It may also ultimately guide the intensity and duration of treatment and minimize the toxicity associated with chronic steroid administration."

"The ability to identify patients who will not respond to traditional treatment and who are at particularly high risk for morbidity and mortality could permit tailored treatment plans, such as additional immunosuppressive treatments for high-risk patients that may be more effective if introduced early," she explained. "Equally important is the identification of low-risk patients who will respond well to treatment. These patients may tolerate a more rapid tapering of steroid regimens to reduce long-term toxicity, infections, and a loss of the graft versus tumor effect."

"It is important to note that these blood biomarkers have been tested in retrospective sets and that they will need to be qualified in prospective samples for future clinical trials," Dr. Paczesny cautioned.

Dr. Roderick Skinner from Great North Children's Hospital's Children's BMT Unit, Royal Victoria Infirmary, Newcastle upon Tyne, UK, told Reuters Health by email, "Due to the potential major importance of the management issue (tailoring initial GVHD treatment), I think that it will be important to replicate the findings in other datasets (with different characteristics in terms of patients, donor types, stem cell sources, HLA-matches, GVHD prophylaxis - since all of these influence acute GVHD risks and outcomes, and also non-relapse mortality). It will be very important to include a pediatric dataset since acute GVHD and particularly non-relapse mortality rates are different in children to those in adults."

Three authors are co-inventors of a patent for use of GVHD biomarkers. The other authors and the accompanying commentators report no relevant financial relationships.


Lancet Haematol 2014


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: