Age-enhanced MAGIC algorithm predicts mortality in pediatric aGVHD: a multicenter study.

Introduction: Acute graft-versus-host disease (aGVHD) is a major contributor to non-relapse mortality (NRM) in pediatric patients undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT). Although the MAGIC algorithm has been validated in adults, its predictive value in children remains insufficiently explored.

Methods: We conducted a prospective multicenter cohort study including 105 Chinese pediatric allo-HSCT recipients diagnosed with aGVHD between May 2019 and August 2023. Endpoints were 6-month NRM, overall survival (OS), and Day-28 treatment response. Multivariable analyses incorporated clinical variables together with the Panel 2 score, hereafter referred to as Panel 2,using Cox regression for NRM/OS and logistic regression for treatment response.

Results: Age ≥12 years (hazard ratio 4.36, 95% CI 1.62-11.75; P=0.003) and a high Panel 2 score (HR 3.09, 95% CI 1.08-8.82; P=0.035) were independent predictors of 6-month NRM and OS. The high-risk (HR) group, defined by the combination of age ≥12 years and a high Panel 2 score, had markedly higher NRM than the low-risk (LR) group (71% vs 12.2%; HR 5.00, 95% CI 1.75-9.56; P=0.001) and significantly worse OS (P<0.001). Panel 2 was also predictive of Day-28 treatment response, with lower CR/PR rates in the high versus low group (62% vs 92%; P<0.001).

Discussion: The Panel 2 score effectively predicted NRM, OS, and treatment response in pediatric aGVHD. Incorporating age ≥12 years further enhanced risk stratification, enabling clear separation between HR and LR groups. These findings support the potential clinical utility of this combined model and warrant validation in larger, international pediatric cohorts.