Introduction: In children undergoing liver transplantation, the in-vivo metabolism of tacrolimus (TAC) is influenced by various factors, including donor CYP3A5 genotype. However, genetic and biological factors are rarely considered in African populations when drug dosing studies are performed. The aim of this study was to determine donor genotype and recipient phenotype characteristics associated with TAC metabolism in children undergoing living donor liver transplantation (LDLT) at our center.
Methods: We conducted an IRB approved study on 22 children undergoing LDLT. Clinical data, biospecimens (blood/liver biopsy) and pharmacy records were collected intra-operatively and for 15 days post-transplant. For all recipients, TAC was administered twice daily, and the dose titrated to achieve trough concentrations between 12-15ng/ml. Pharmacokinetic (PK) outcome measures were: daily TAC trough levels,  TAC dose (mg/dose), mean TAC dose, TAC concentration dose ratio adjusted for patient weight (CDR). Clinical outcome measures included acute cellular rejection episodes (ACR) in the first 90 days post-transplant. These outcomes were compared between 3 donor CYP3A5 genotype groups: *1/*1 homozygous enzyme expressors; *1/*n heterozygote enzyme expressors; and *3/*3 homozygous enzyme non-expressors, using one-way ANOVA or Kruskal-Wallis tests. Associations between the recipient age, body weight, graft weight recipient weight ratio (GRWR) and PK outcome measures were determined by Spearman’s rank correlation coefficient. 
Results: Of 22 transplants performed, donor ethnicities were black (n=13); white (n=4), and other (n=5). All donors of black ethnicity had the CYP3A5*1*1 /or *1/n genotype, with CYP3A5*3/*3/ or *1/n genotypes in the remaining ethnic groups. CYP3A5*1*1 recipients required significantly higher mean (sd) doses of TAC 3.6mg (1.2) compared with CYP3A5*3/*3 recipients who required 1.7mg (0.6) (p=0.039). The median (IQR) CDR was significantly lower in CYP3A5*1 recipients (*1/*1: 32 (25-37); *1*n: 57 (21-110)) compared with the 3/*3 genotype 183 (88-247) (p=0.016) (Figure 1). There was no significant difference between CYP3A5 genotypes and ACR (p=0.23). The mean CDR was significantly correlated with recipient age (rs 0.50), body weight (rs 0.53), and GRWR (rs -0.53). 
Conclusion: We have shown that despite daily therapeutic drug monitoring (TDM), children receiving donor grafts with good enzyme activity (CYP3A5*1) had relatively low TAC levels and required higher doses of TAC compared to CYP3A5*3 donor genotypes. We confirmed the impact of recipient age and weight on TAC PK as well the influence of a larger graft on metabolic capacity. A precision medicine approach to TAC dosing algorithms could be guided by individual genetic (CYP3A5 genotype) and clinical profiles (age, weight and GRWR) in children undergoing LDLT to achieve target concentrations thereby reducing the risks of either over- or under immunosuppression, particularly in understudied populations.