Predictors of grade III-IV acute GVHD.

(A) Effect of T-cell depletion. Patients who received a T-cell–depleted transplant had an estimated day-100 probability of grade III-IV acute GVHD of 8% versus 50% if no T-cell depletion was performed (P < .001, log-rank test). (B) Effect of the pretransplant hepatic transaminase values before starting the conditioning regimen. Patients with elevated serum transaminase greater than twice the upper normal values had an estimated day-100 probability of grade III-IV acute GVHD of 64% versus 26% if it was below that level (P = .01, log-rank test). (C) Effect of the limb malformations. Patients with malformations of the limbs had an estimated day-100 probability of grade III-IV acute GVHD of 55% versus 23% if no limb malformation was observed (P < .02, log-rank test). (D) Effect of the urogenital and kidney malformations. Patients with urogenital and/or kidney malformations had an estimated day-100 probability of grade III-IV acute GVHD of 50% versus 26% if such malformations were not present (P = .04, time-dependent Cox proportional hazards model).