Tacrolimus is a key immunosuppressant used in solid organ transplantation. However, its pharmacokinetics exhibit significant interpatient variability and it has a narrow therapeutic window, which makes predicting dosage and plasma concentrations difficult. It is primarily metabolized by CYP3A4/5 and polymorphisms in the CYP3A4/5 gene, P-glycoprotein and inflammatory pathways can all interfere with this process. Therefore, therapeutic drug monitoring is necessary to identify the various factors affecting its metabolism ^[4]. Current hypotheses suggest that EBV infection may elevate inflammatory cytokines, downregulate CYP450 enzyme expression and consequently alter tacrolimus trough concentrations ^[5]. Based on this hypothesis, the present study investigated whether EBV serological status and viral load influence the pharmacokinetic characteristics of tacrolimus in pediatric liver transplant recipients.

This was a single-center, retrospective study. All enrolled patients were required to undergo monthly EBV titer and tacrolimus TDM testing during the first year following liver transplantation, with both tests being performed during the same week. After one year, the frequency of these tests was reduced to once every three months. As the study focused on changes in trough concentrations, samples with concentrations exceeding 15 μg /mL or falling below 1 μg /mL due to non-compliance with medical instructions were excluded.

First, the study conducted a cross-sectional analysis of 34 pediatric outpatients to determine the association between EBV serological status and weight-adjusted tacrolimus concentration-to-dose ratio (C/D ratio), thereby minimizing the effects of concomitant medications on tacrolimus pharmacokinetics. The subsequent longitudinal study focused on 36 pre-transplant EBV-negative patients aged 0–2 years. These patients were divided into two groups based on subsequent EBV test results (positive vs. negative) and the C/D ratio was ultimately analyzed.

A cross-sectional study of 34 participants showed that women constituted the majority (66%) and that the primary cause of their condition was biliary atresia. Four of the participants were taking medications that affect tacrolimus metabolism, such as itraconazole, flucloxacillin and nifedipine, during treatment.

An analysis of EBV infection in pediatric transplant recipients revealed that 11 patients remained EBV-negative throughout treatment, 14 were EBV-positive throughout treatment, and 11 became EBV-positive during treatment. Notably, EBV-negative patients were predominantly male, while EBV-positive patients and those who became EBV-positive were predominantly female, indicating a significant gender difference (p=0.028).

To determine the relationship between EBV serological status and the weight-adjusted tacrolimus C/D ratio, the study integrated two datasets — one cross-sectional and one longitudinal — and made comparisons from both between- and within-patient perspectives. Intergroup comparisons revealed only minor differences in the median and interquartile range of the C/D ratio between EBV-negative and EBV-positive subjects (as shown in Figure 1), with no clear trend in the data distribution. Intrapatient curves also indicated that the C/D ratio did not regularly increase or decrease following seroconversion from EBV-negative to EBV-positive. Analysis using a linear mixed model yielded a p-value of 0.85, confirming that EBV serostatus was not statistically associated with tacrolimus metabolic levels (see Figure 2). Furthermore, the C/D ratio exhibited high inter-individual variability and spanned a very wide range across all subjects, indicating that EBV is not a key factor contributing to fluctuations in tacrolimus concentrations.

Figure 1: The interpatient association between EBV serostatus and the tacrolimus weight-adjusted C/D ratio in the longitudinal database. The figure includes patients that remained EBV-negative and patients that remained EBV-positive during longitudinal follow-up. C/D concentration-to-dose, EBV Epstein–Barr virus, EBV- EBV-negative, EBV+ EBV-positive, n number of patients.

Figure 2, The intrapatient association between EBV serostatus and the tacrolimus weight-adjusted C/D ratio in the cross-sectional and longitudinal database. Recipients who converted from EBV-negative to EBV-positive were included. No significant difference was observed between EBV-negative and EBV-positive recipients. C/D concentration-to-dose, EBV Epstein–Barr virus, EBV− EBV-negative, EBV+ EBV-positive, n number of patients.

To investigate the relationship between EBV viral load and the C/D ratio further, the study divided EBV-positive patients into four groups according to their viral load. The median C/D ratio remained stable across all groups, with no linear trend of simultaneous increase or decrease as viral load increased (see Figure 3). The statistical test result was P = 0.85, indicating no significant correlation between the two variables. Although there were a few outliers with extremely high values in the low viral load group, these fluctuations were due to inherent differences between patients rather than changes in viral load.

Figure 3，The intrapatient association between EBV viral load and the tacrolimus weight-adjusted C/D ratio in the cross-sectional and longitudinal databases. No significant difference was observed between the groups. C/D concentration-to-dose, EBV Epstein–Barr virus, n number of patients.

Based on the results of the above analysis, neither EBV serostatus (positive or negative) nor different levels of EBV viral load had a significant impact on the adjusted tacrolimus concentration-to-dose ratio in pediatric liver transplant recipients when used as observational variables. Collectively, the study data confirm that the initial hypothesis is not supported, and that there is no clinical need to adjust tacrolimus dosing based on EBV serostatus or changes in viral load. Substantial interpatient variability in drug concentrations is more likely to be driven by factors such as genetics, concomitant medications and graft-related parameters. The overall statistical results are shown in Table 1.

Table1 Patient characteristics of the various study cohorts in our study