Abstract
BACKGROUND: Calcineurin inhibitors balance graft-versus-host disease (GvHD) prevention against graft-versus-leukemia (GvL) effects after allogeneic hematopoietic cell transplantation (allo-SCT). Whether the magnitude of early tacrolimus exposure influences relapse or survival remains unclear. METHODS: We retrospectively studied 122 adults with AML (n = 80) or MDS (n = 42) undergoing 10/10 matched-unrelated donor (MUD) allo-SCT from 2014 to 2021. All received rabbit anti-thymocyte globulin (ATG), tacrolimus and mycophenolate mofetil (MMF) for GvHD Prophylaxis. Tacrolimus exposure was quantified as area under the concentration-time curve over days 1-30 (AUC30) and 1-100 (AUC100). Primary endpoints were overall survival (OS) and event-free survival (EFS); secondary endpoints were cumulative incidence of relapse (CIR), non-relapse mortality (NRM), and GvHD. Analyses included median splits, quartiles/deciles, competing-risks models, Cox regression, and smoothed 5-year KM estimates across continuous AUC30. Patients with relapse/death before day + 30/+100 were excluded from the respective AUC analyses. RESULTS: Median follow-up was 61 months; 5-year OS and EFS were 68% and 43%. Neither AUC30 nor AUC100 (median split or quartiles) was associated with OS, EFS, CIR, or NRM. In multivariable models, tacrolimus exposure was not significant; worse outcomes correlated with older age and female donor for OS, and with AML diagnosis and no CR at allo-SCT for EFS. Smoothed survival across continuous AUC30 showed no meaningful gradient. Any-grade and severe aGvHD, and overall/grade ≥ III chronic GvHD, were comparable across exposure groups. CONCLUSIONS: Within a homogeneous MUD/ATG/MMF setting, early tacrolimus exposure did not independently predict relapse, NRM, EFS, or OS. Patient and donor factors outweighed tacrolimus AUC. Findings argue against AUC-targeting to improve outcomes; risk-adapted tapering and MRD-guided interventions may offer greater leverage.