Decoding Graft-versus-Host Disease: Can We Predict Patient Outcomes?
"New research sheds light on biomarkers that could predict survival rates in acute graft-versus-host disease (GVHD) patients after reduced-intensity conditioning."
For over a decade, the study and analysis of biomarkers to diagnose and predict disease outcomes have seen a surge. These biomarkers, which can be any clinical parameter or protein linked to an individual's health status, are assessed using both new and established platforms to visualize protein and peptide expression in tissues and body fluids. Much of the research focuses on discovering new protein or peptide biomarkers, clusters, or patterns in body fluids like blood and urine.
Allogeneic hematopoietic stem-cell transplantation (HSCT) serves as a curative treatment for patients battling hematological malignancies or hematopoietic failure syndromes. However, its use is often hindered by severe complications, notably acute and chronic graft-versus-host disease (GVHD). Consequently, numerous research teams have been dedicated to identifying biomarkers capable of predicting the onset of acute GVHD, offering prognostic insights for detecting steroid resistance, and ultimately, forecasting mortality post-GVHD. A notable advancement in this area is the use of urine proteomics, employing capillary electrophoresis coupled online with mass spectrometry, which currently stands as the sole model for predicting acute GVHD development.
The Endothelial Activation and Stress Index (EASIX), which combines routinely monitored clinical biomarkers post-HSCT, offers a practical approach. Based on lactate dehydrogenase, creatinine, and platelet count, EASIX predicts mortality in acute GVHD patients after reduced-intensity conditioning. Its straightforward formula—lactate dehydrogenase (U/L) × creatinine (mg/dL)/thrombocytes (10⁹ cells per L)—makes identifying patients unlikely to respond to steroid therapy remarkably simple.
Changing Rates and Outcomes
At Fred Hutchinson Cancer Center, prospective tracking using the National Institutes of Health (NIH) Consensus Criteria began in 2002, and chronic GVHD requiring systemic immunosuppressive therapy has steadily declined since 2005. A 2024 report also describes declining acute GVHD probabilities over the last decade, particularly among children and recipients transplanted for nonmalignant diseases; across three time windows, two-year survival estimates improved, while nonrelapse mortality trended downward. These trends indicate change over time, but the supplied sources do not provide a single current incidence figure for all patients.
Evolving Prevention and Clinical Guidance
The NIH Consensus Development Project held its third conference on chronic GVHD clinical-trial criteria in 2020 to identify gaps in understanding, prevention, and treatment and to make actionable recommendations. Historically, cord-blood and matched-sibling transplantation have been associated with the lowest GVHD rates, while newer prevention methods have modified practice. Clinical approaches to prophylaxis and management have also changed with regulatory approvals of new drugs in the three years before a 2024 review. Despite advances in prophylaxis and decreasing severe acute GVHD incidence, important clinical questions remain.
From Secondary Disease to GVHD
In early transplantation experiments, high-dose total-body irradiation (TBI) caused mice to die from primary disease or bone-marrow aplasia. Intravenous allogeneic bone-marrow cells delayed mortality but did not prevent it; the later deaths were initially called “secondary disease.” That condition was subsequently renamed graft-versus-host disease (GVHD). The historical account describes donor T lymphocytes as attacking recipient tissues including skin, intestinal mucosa, bile ducts, and lymph nodes.
EASIX-GVHD: A Promising Tool with Limitations
The ability to predict mortality in acute GVHD patients has garnered significant attention in recent years. Protein biomarkers, assessed at the onset of acute GVHD using ELISA, have been extensively published. While initial studies suggested the necessity of measuring six biomarkers for accurate prediction, advancements have refined this to two key markers: suppression of tumorigenicity 2 (ST2) and serum or plasma Reg3alpha. While these biomarkers effectively predict mortality post-acute GVHD onset and overall HSCT outcomes, the costs associated with the platform used to acquire this information can be substantial. In contrast, EASIX-GVHD, as described by Thomas Luft and colleagues, utilizes routinely measured parameters in allogeneic HSCT patients.
- Variations in Conditioning Intensities: Reduced-intensity conditioning varies considerably, which can lead to substantial variations in platelet counts, potentially affecting the accuracy of survival predictions.
- Limited Scope of Parameters: The hazard ratio of overall survival is only 1.2 in both univariable and multivariable analyses, suggesting that the parameters building up the EASIX-GVHD score cannot account for all of the pathophysiological changes occurring during acute GVHD.
Research on Preemption and Tissue-Resident Cells
In a randomized phase II/III trial, short-term preemptive prednisolone did not significantly change the incidence or severity of acute GVHD (hazard ratio 1.69, 95% confidence interval 0.66–4.32; P = 0.27). A 2026 Nature research summary reports that donor tissue-resident CD4 T cells contribute to chronic GVHD, while how these cells are maintained and their pathogenic mechanism remain undefined. A 2026 review also surveys classical acute GVHD, chronic GVHD, and overlap syndromes as defined by NIH Consensus Criteria.
Donor Matching and GVHD Risk
A review reports GVHD in about 30–50% of recipients of matched-related donor grafts and 70% of recipients of matched-unrelated donor grafts. It identifies the degree of HLA disparity between donor and recipient as an established independent risk factor for GVHD development. These figures compare the two donor groups reported in that review; the supplied sources do not establish a broader comparison across transplant approaches.
Future Directions: Refining Prediction Models for GVHD Outcomes
Luft and colleagues' work adds to the existing scientific literature on biomarker development for prediction of death after GVHD with a simple tool that can be easily applied to any patient cohort.
Changing Care and Ocular Complications
A 2024 review describes GVHD as a major contributor to morbidity and mortality after allogeneic hematopoietic stem-cell transplantation and notes regulatory approval of new drugs alongside substantial changes in prophylaxis and management approaches in the preceding three years. An expert review describes ocular GVHD as a fibroinflammatory complication after transplantation, with severe, cicatrizing dry eye as its primary symptom. The review says this condition can progress to an irreversible state; another supplied source characterizes GVHD-associated dry eye as involving extensive ocular-surface inflammatory destruction and pain and visual impairment.
Market and Epidemiology Outlook
One market report values the global GVHD treatment market at USD 2.8 billion in 2023 and projects a compound annual growth rate of 8.2% from 2024 to 2030. The report attributes the projected expansion to an aging population and increased occurrence of nephrological, leukemia, myeloma, and lymphoma cancers. Separately, a 2026 epidemiology forecast examines prevalence trends by age group, gender, and disease type, but the supplied snippet gives no prevalence estimates.
Expansion of the present cohorts might lead to improved insights into the importance of conditioning regimens, and whether EASIX-GVHD retains its predictive value in patients receiving different reduced-intensity conditioning protocols remains to be shown in future studies.
Further studies with larger cohorts, especially comparing the biomarkers described for prediction of acute GVHD and outcome with the biomarkers for prediction of outcome after acute GVHD, preferably done in the centres that first published the analyses of the biomarkers, would be interesting and are well overdue. EASIX-GVHD could be included in these analyses and the predictive value of the markers could be analysed in a large multicentre trial.