Outcome predictions for patients with graft-versus-host disease

From the Hematopoietic Cell Transplantation Program, Clinical Research Division

The risk of developing graft-versus-host disease (GVHD) looms over the head of allogenic hematopoietic stem cell transplant recipients. GVHD is a complex disease state, whereby cells from the transplant donor recognize healthy cells of the transplant recipient as a foreign entity that must be cleared from circulation, attacking, and killing them. This causes immense risk to the recipient, leading to an increased risk of mortality. When first-line treatment with steroids is not effective, second-line treatment is needed. The outcomes for this patient group after second-line treatment are poor, with recent studies describing survival rates of approximately 50% 6-months after beginning second-line treatments. Dr. Phuong Vo, Dr. Paul Martin, both members of Fred Hutch’s Clinical Research Division, and colleagues sought to investigate the predictability of outcomes after second-line treatment for GVHD. They led a retrospective study examining reported outcomes from patients who underwent treatment for acute GVHD after transplantation by assessing overall survival and failure-free survival rates amongst study participants. Here, they designed a predictive statistical model that identified a set of readily available clinical variables that were strongly associated with mortality risk for this patient population. Explaining the basis of their study, recently published in Blood Advances, Dr. Vo said “GVHD requiring second-line treatment represents a highly morbid complication of allogenic hematopoietic cell transplantation. Recent studies have defined short-term outcomes after second-line treatment for acute GVHD, but longer-term outcomes have not been well defined. The results of this study are important because they provide historical benchmarks for the longer-term outcomes of overall survival and failure free survival that would be useful in assessing results of future single-arm and controlled trials testing new agents for second-line treatment of acute GVHD.”

Participants were selected retrospectively if they had come to Fred Hutchinson Cancer Center between 2006 and 2018 for hematopoietic cell transplantation and needed second-line treatment of acute GVHD.  Survival without recurrent malignancy or third-line systemic treatment for acute GVHD (i.e., failure-free survival) at 6 months after the onset of second-line treatment was estimated to be 42% and overall survival was 59%, a slightly higher figure than reported in other studies, which the authors attributed to the higher proportion of pediatric patients in their study.

Age, abdominal pain, and serum bilirubin and albumin concentrations can predict mortality outcomes for patients with acute graft-versus-host disease.
Age, abdominal pain, and serum bilirubin and albumin concentrations can predict mortality outcomes for patients with acute graft-versus-host disease. Image provided by Dr. Vo.

Next, the authors performed statistical analysis to determine potential clinical variables that could be associated with mortality risk. Ruling out factors such as transplant number and fever, they observed that older patient age, lower serum albumin, higher serum bilirubin, and stage 4 gut/abdominal pain were associated with higher overall mortality risk. Except for patient age, these same factors were also associated with an increased risk of treatment failure (i.e., death, recurrent malignancy or third-line systemic treatment). Notably, when the authors adjusted their model for these risk factors, none of the agents used for second-line treatment was demonstrably better or worse than others with respect to overall mortality or treatment failure.

By using these same clinical variables, the authors developed mathematical models that predicted the probabilities of death at 6 and 12 months after starting second-line treatment. “Our identified risk factors will also help evaluate whether differences between trials or study arms might be attributable to patient selection rather than the effects of the investigational product. Furthermore, the models for 6- and 12-month mortality can be used to generate population-specific benchmarks in future studies and to provide valuable prognostic information in counseling patients who require second-line treatment for acute GVHD,” highlighted Dr. Vo.  

The authors anticipate that the findings from this study will help inform future clinical trials of the importance of including overall survival as an endpoint for this patient population. “Historically, overall survival has not served as the primary endpoint in trials of treatment for steroid-refractory or -dependent acute GVHD, primarily because many other variables unrelated to acute GVHD might affect survival of these patients and complicate the analysis and interpretation. Our results suggest that survival can be predicted reasonably well with a limited number of clinical variables and raise the question if the future studies should hold improved overall survival as a goal to reach,” discussed Dr. Vo. Future work for their group is focused on combining the models derived in this study with ongoing biomarker research to further understand and predict mortality outcomes for patients with acute GVHD. “It would be interesting to utilize these models we developed in future studies to compare the accuracy and utility of the clinical risk factor approach versus the biomarker-based approach (e.g., combination of REG3α and ST2 with or without TNFR1) and to evaluate a combination of the two approaches for calibrated prediction of survival after second-line systemic treatment for acute GVHD,” summarized Dr. Vo.

This work was funded by the National Institutes of Health.

Fred Hutch/University of Washington/Seattle Children's Cancer Consortium members Dr. Phuong Vo, Dr. Ted A. Gooley, Dr. Paul A. Carpenter, Dr. Mohamed L. Sorror and Dr. Paul J. Martin contributed to this work.

Vo P, Gooley TA, Carpenter PA, Sorror ML, MacMillan ML, DeFor TE, Martin PJ. Prediction of outcomes after second-line treatment for acute graft-versus-host disease. Blood Adv. 2022 Jun 14;6(11):3220-3229. doi: 10.1182/bloodadvances.2021006220. PMID: 35235948; PMCID: PMC9198915.