Abstract
BACKGROUND: Tacrolimus therapy is standard of care for immunosuppression after lung transplantation. However, tacrolimus exposure variability during the early postoperative period may contribute to poor outcomes in this population. Few studies have examined tacrolimus pharmacokinetics (PK) during this high-risk period. METHODS: We conducted a retrospective pharmacokinetic study in lung transplant recipients at the University of Pennsylvania who were enrolled in the Lung Transplant Outcomes Group cohort. We used nonlinear mixed-effects regression to derive a population PK model in 270 patients and examined validity in a separate cohort of 114 patients. Covariates were examined with univariate analysis and a multivariable model was developed using forward and backward stepwise selection. The performance of the final model in the validation cohort was examined with calculation of prediction error (PE). RESULTS: We developed a 1-compartment base model with a fixed rate absorption constant. Covariates improving model fit were postoperative day, hematocrit, transplant type, CYP3A5 genotype, weight, and exposure to cytochrome p450 enzyme (CYP) inhibitor drugs. The strongest predictor of tacrolimus clearance was postoperative day, with median predicted clearance increasing more than 3-fold over the 14-day study period. In the validation cohort, the final model showed a mean PE of 36.4% (95% confidence interval 30.8%-41.9%) and a median PE of 7.2% (interquartile range -29.3% to 70.53%). CONCLUSIONS: Tacrolimus clearance is highly dynamic during the early postlung transplant period. Population PK models that include lung-transplant-specific covariates may enable precision dosing algorithms that account for this highly dynamic clearance. Future multicenters studies including a broader set of covariates are warranted.