Pharmacokinetic and Pharmacodynamic Variability Predict Late Events After Paediatric Liver Transplantation:...

Cross-sectional tacrolimus concentrations and liver biochemistry may not reflect longitudinal instability after paediatric liver transplantation.

To derive a risk score based on pharmacokinetic and biochemical variability at one centre and validate it at an independent centre.

This retrospective two-centre landmark study included 93 recipients, predominantly infants with biliary atresia (83.9%) undergoing primary liver transplantation, the majority receiving living-donor grafts (88.2%), and 86 recipients in the validation cohort (76.7% biliary atresia; 45.3% living-donor). Variability in tacrolimus trough concentrations, aspartate aminotransferase and total bilirubin was calculated from measurements obtained 6-12 months after transplantation. Patients with major complications during this period were excluded. Follow-up began at 12 months. The primary outcome was the first major late complication or death without a preceding qualifying complication. A three-point score was derived and applied unchanged to the external cohort.

Nineteen derivation-cohort patients and 12 validation-cohort patients experienced late events. The score assigned one point each for tacrolimus variability above 25%, aspartate aminotransferase variability above 50% and total bilirubin variability above 50%. Scores of 2-3 identified higher-risk patients in the derivation cohort, with a hazard ratio of 4.48, a 95% confidence interval of 1.79-11.17 and a concordance index of 0.748. Areas under the curve at 1, 3 and 5 years were 0.883, 0.802 and 0.740. In the external cohort, the hazard ratio was 8.90, with a 95% confidence interval of 2.66-29.81 and corresponding areas under the curve of 0.753, 0.775 and 0.789.

Integrating the longitudinal variability of tacrolimus exposure and liver biochemistry provides a practical, externally validated, non-invasive tool for paediatric liver transplantation survivors.