A Biomarker-Integrated Day 7 Response Reliably Predicts Outcomes in Minnesota Standard Risk Acute GVHD

In 790 Minnesota standard risk acute GVHD patients, a biomarker-integrated day 7 response model predicts 6-month non-relapse mortality with higher discrimination than day 7 clinical response alone (AUC 0.76 vs 0.63).

• Why it matters: Day 7 clinical response alone is limited for early risk sorting, because many patients without early improvement later respond without escalation. Earlier, more objective stratification is needed for treatment decisions and trial endpoints.
• What they did: Using a MAGIC observational cohort split into training (n=294) and validation (n=308), plus an external interventional cohort (n=188), a classification and regression tree combined day 7 clinical response with Ann Arbor biomarker scores to predict 6-month NRM.
• The result: The integrated model reclassified about 20% of patients and reproduced across validation cohorts, improving 6-month NRM prediction (P=.002) and matching day 28 clinical response performance (AUC 0.70). Among day 7 non-responders, biomarker status distinguished favorable slow responders and stratified outcomes even after second-line therapy.

AI summary of the abstract · may contain errors · not medical advice · read the paper.