Impact of dose-adjusted tacrolimus exposure phenotype on outcomes in kidney transplantation: a large-scale multicenter cohort study

Abstract

Introduction: Dose-adjusted tacrolimus exposure phenotype, expressed as the trough concentration-to-dose ratio (C0/Dose), may influence post-transplant outcomes, but supporting evidence remains limited. Methods: We evaluated the association between tacrolimus exposure phenotype and short- and long-term allograft and patient outcomes in a multicenter retrospective cohort study using clinical data warehouse records from five tertiary transplant centers in South Korea (2005-2020). Patients were classified into fast- or slow-dose-adjusted exposure phenotypes using trajectory clustering. Primary outcomes included 1-year endpoints (graft failure, biopsy-proven acute rejection, de novo donor-specific antibodies, severe infection, cardiovascular events, malignancy, and death). Secondary outcomes were assessed using the same endpoints for 2-6 years. Inverse probability of treatment weighting (IPTW) and weighted Cox proportional hazards models were used to adjust for confounders. Results: Among 5,965 recipients, 4,595 (77.0%) were classified as fast-exposure phenotype and 1,370 (23.0%) as slow-exposure phenotypes. Slow-exposure phenotypes received lower tacrolimus doses (time-weighted median 2.1 vs 4.6 mg/day, P < 0.001) but had higher trough levels (6.7 vs 6.1 ng/mL, P < 0.001), higher variability (27.8 vs 26.4%, P < 0.001), and greater time above the therapeutic range (19.7 vs 11.6%, P < 0.001). After IPTW, 1-year efficacy outcomes (biopsy-proven acute rejection, graft failure, and de novo donor-specific antibody development) did not differ significantly between groups. During years 2-6, slow-exposure phenotypes had higher risks of all-cause mortality (adjusted HR 1.616, 95% CI, 1.069-2.442), serious infection requiring hospitalization (adjusted HR 1.537, 95% CI, 1.166-2.026), and malignancy (adjusted HR 1.628, 95% CI, 1.077-2.460). Discussion: A slow tacrolimus exposure phenotype (high C0/Dose) may serve as a risk marker of cumulative overexposure, associated with increased late mortality, serious infection, and malignancy. Phenotype-informed longitudinal exposure monitoring may improve long-term risk stratification, although whether targeted dose adjustment improves outcomes requires prospective interventional studies.