The MAGIC algorithm probability (MAP)-guided preemptive therapy of acute graft versus host disease with methylprednisolone: A randomized controlled trial.

ABSTRACT

Acute graft versus host disease (aGvHD) is a severe complication that arises in patients undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT) and remains the primary cause of nonrelapse mortality (NRM). The MAGIC algorithm probability (MAP) has been proposed to identify patients at intermediate and high risk of developing aGvHD. The levels of suppression of tumorigenicity 2 (ST2) and regenerating islet-derived 3α (Reg3α) were assessed, and MAP was calculated on days 7, 14, 21, and 28 after allo-HSCT. Based on the MAP results, patients were classified into low-, intermediate-, or high-risk groups for the development of aGvHD. Random assignment was performed to allocate intermediate- or high-risk patients to receive preemptive therapy with methylprednisolone or not. The 100-day cumulative incidences of grade 2 or higher (35.5% ± 8.6%) and grade 3 or higher (12.9% ± 6.0%) aGvHD in the methylprednisolone group were significantly lower than those in the control group (66.7% ± 7.9%, p = .01; 42.9% ± 8.4%, p = .01), and similar to those observed in the low-risk group (31.7% ± 7.3%, p = .75; 2.4% ± 2.4%, p = .08). The 6-month cumulative incidences of NRM were 14.1% ± 6.6%, 22.7% ± 7.1%, and 2.4% ± 2.4% in the methylprednisolone, control, and low-risk groups, respectively, with no significant difference between the methylprednisolone and control groups (p = .29). Methylprednisolone did not increase infections (p = .34). The 100-day cumulative incidences of cytomegalovirus (CMV) reactivation were 67.7% ± 8.4%, 65.6% ± 8.4%, and 46.3% ± 7.8% (p = .08), and those of grade 2 or higher hemorrhagic cystitis were 29.0% ± 8.2%, 45.2% ± 8.9% and 22.0% ± 6.5% (p = .11) in the methylprednisolone, control, and low-risk groups, respectively. MAP-guided preemptive therapy for aGvHD is promising. The long-term efficacy and safety remain to be investigated.