Test predicts response to GVHD treatment

A newly developed algorithm could enable risk-adapted therapy for graft-vs-host disease (GVHD), researchers have reported in Lancet Haematology.

The problem with treating GVHD, the team said, is that the severity of symptoms at disease onset does not accurately define risk.

Patients with low-risk GVHD are often overtreated and experience significant side effects, while patients with high-risk GVHD can be undertreated and see their GVHD progress.

“Our goal is to provide the right treatment for each patient,” said study author James L. M. Ferrara, MD, DSc, of the Icahn School of Medicine at Mount Sinai.

“We hope to identify those patients at higher risk and design an aggressive intervention while tailoring a less-aggressive approach for those with low-risk.”

With that goal in mind, Dr Ferrara and his colleagues developed and tested a scoring system for GVHD. They collected plasma from 492 patients newly diagnosed with varying grades of GVHD and randomly assigned them to training (n=328) and test (n=164) sets.

The team used the concentrations of 3 validated biomarkers—TNFR1, ST2, and Reg3α—to create an algorithm that calculated the probability of non-relapse mortality (NRM) 6 months after GVHD onset for patients in the training set.

The researchers ranked the probabilities and identified thresholds that created 3 NRM scores. They then tested the algorithm in the test set of patients and a validation cohort of 300 additional patients who were enrolled on trials of GVHD treatment.

Results showed the algorithm works. The cumulative incidence of 6-month NRM significantly increased as the Ann Arbor GVHD score increased, and the response to primary GVHD treatment within 28 days decreased as the GVHD score increased.

In the validation set, the incidence of NRM was 8% for score 1, 27% for score 2, and 46% for score 3 (P<0.0001). Treatment response measured 86% for score 1, 67% for score 2, and 46% for score 3 (P<0.0001).

“This new scoring system will help identify patient who may not respond to standard treatments and may require an experimental and more aggressive approach,” Dr Ferrara said. “And it will also help guide treatment for patients with lower-risk GVHD who may be overtreated. This will allow us to personalize treatment at the onset of the disease. Future algorithms will prove increasingly useful to develop precision medicine for all [stem cell transplant] patients.”

To capitalize on this discovery, Dr Ferrara created the Mount Sinai Acute GVHD International Consortium (MAGIC), which consists of a group of 10 stem cell transplant centers in the US and Europe that will collaborate to use this new scoring system to test new treatments for acute GVHD.

Dr Ferrara and his colleagues have also written a protocol to treat high-risk GVHD that has been approved by the US Food and Drug Administration.

A newly developed algorithm could enable risk-adapted therapy for graft-vs-host disease (GVHD), researchers have reported in Lancet Haematology.

The problem with treating GVHD, the team said, is that the severity of symptoms at disease onset does not accurately define risk.

Patients with low-risk GVHD are often overtreated and experience significant side effects, while patients with high-risk GVHD can be undertreated and see their GVHD progress.

“Our goal is to provide the right treatment for each patient,” said study author James L. M. Ferrara, MD, DSc, of the Icahn School of Medicine at Mount Sinai.

“We hope to identify those patients at higher risk and design an aggressive intervention while tailoring a less-aggressive approach for those with low-risk.”

With that goal in mind, Dr Ferrara and his colleagues developed and tested a scoring system for GVHD. They collected plasma from 492 patients newly diagnosed with varying grades of GVHD and randomly assigned them to training (n=328) and test (n=164) sets.

The team used the concentrations of 3 validated biomarkers—TNFR1, ST2, and Reg3α—to create an algorithm that calculated the probability of non-relapse mortality (NRM) 6 months after GVHD onset for patients in the training set.

The researchers ranked the probabilities and identified thresholds that created 3 NRM scores. They then tested the algorithm in the test set of patients and a validation cohort of 300 additional patients who were enrolled on trials of GVHD treatment.

Results showed the algorithm works. The cumulative incidence of 6-month NRM significantly increased as the Ann Arbor GVHD score increased, and the response to primary GVHD treatment within 28 days decreased as the GVHD score increased.

In the validation set, the incidence of NRM was 8% for score 1, 27% for score 2, and 46% for score 3 (P<0.0001). Treatment response measured 86% for score 1, 67% for score 2, and 46% for score 3 (P<0.0001).

“This new scoring system will help identify patient who may not respond to standard treatments and may require an experimental and more aggressive approach,” Dr Ferrara said. “And it will also help guide treatment for patients with lower-risk GVHD who may be overtreated. This will allow us to personalize treatment at the onset of the disease. Future algorithms will prove increasingly useful to develop precision medicine for all [stem cell transplant] patients.”

To capitalize on this discovery, Dr Ferrara created the Mount Sinai Acute GVHD International Consortium (MAGIC), which consists of a group of 10 stem cell transplant centers in the US and Europe that will collaborate to use this new scoring system to test new treatments for acute GVHD.

Dr Ferrara and his colleagues have also written a protocol to treat high-risk GVHD that has been approved by the US Food and Drug Administration.

A newly developed algorithm could enable risk-adapted therapy for graft-vs-host disease (GVHD), researchers have reported in Lancet Haematology.

The problem with treating GVHD, the team said, is that the severity of symptoms at disease onset does not accurately define risk.

Patients with low-risk GVHD are often overtreated and experience significant side effects, while patients with high-risk GVHD can be undertreated and see their GVHD progress.

“Our goal is to provide the right treatment for each patient,” said study author James L. M. Ferrara, MD, DSc, of the Icahn School of Medicine at Mount Sinai.

“We hope to identify those patients at higher risk and design an aggressive intervention while tailoring a less-aggressive approach for those with low-risk.”

With that goal in mind, Dr Ferrara and his colleagues developed and tested a scoring system for GVHD. They collected plasma from 492 patients newly diagnosed with varying grades of GVHD and randomly assigned them to training (n=328) and test (n=164) sets.

The team used the concentrations of 3 validated biomarkers—TNFR1, ST2, and Reg3α—to create an algorithm that calculated the probability of non-relapse mortality (NRM) 6 months after GVHD onset for patients in the training set.

The researchers ranked the probabilities and identified thresholds that created 3 NRM scores. They then tested the algorithm in the test set of patients and a validation cohort of 300 additional patients who were enrolled on trials of GVHD treatment.

Results showed the algorithm works. The cumulative incidence of 6-month NRM significantly increased as the Ann Arbor GVHD score increased, and the response to primary GVHD treatment within 28 days decreased as the GVHD score increased.

In the validation set, the incidence of NRM was 8% for score 1, 27% for score 2, and 46% for score 3 (P<0.0001). Treatment response measured 86% for score 1, 67% for score 2, and 46% for score 3 (P<0.0001).

“This new scoring system will help identify patient who may not respond to standard treatments and may require an experimental and more aggressive approach,” Dr Ferrara said. “And it will also help guide treatment for patients with lower-risk GVHD who may be overtreated. This will allow us to personalize treatment at the onset of the disease. Future algorithms will prove increasingly useful to develop precision medicine for all [stem cell transplant] patients.”

To capitalize on this discovery, Dr Ferrara created the Mount Sinai Acute GVHD International Consortium (MAGIC), which consists of a group of 10 stem cell transplant centers in the US and Europe that will collaborate to use this new scoring system to test new treatments for acute GVHD.

Dr Ferrara and his colleagues have also written a protocol to treat high-risk GVHD that has been approved by the US Food and Drug Administration.