Background and Importance Cyclosporine is an immunosuppressive drug with complex pharmacokinetics, a narrow therapeutic interval and dose-related adverse effects (nephrotoxicity, hepatotoxicity, and neurotoxicity).

Amiodarone, verapamil and macrolides increase cyclosporine serum concentrations (CSC), whereas other drugs such as phenytoin, carbamazepine and rifampin decrease CSC.

Therefore, therapeutic drug monitoring of cyclosporine is of great importance in routine clinical practice.

Aim and Objectives

• Design a population pharmacokinetic model of cyclosporine.

• Analyse the influence of the recorded covariates.

Material and Methods Retrospective observational study that included patients hospitalised at Severo Ochoa University Hospital and treated with cyclosporine between January 2016 and April 2022. Patients hospitalised in the ICU and outpatients were excluded.

Data recorded date, time and value of the CSC, route of administration, doses administered, sex, age, weight, haematocrit, albumin, serum creatinine and concomitant treatment.

We tested the one- and two-compartmental models with four estimations: first order, first order with interaction, first order conditional and first order conditional with interaction. The influence of the recorded covariates was evaluated, selecting those that showed a statistically significant reduction in the objective function (OFV).

Patients included 29 patients, aged 65 years-old (28–92), 66,7% female. Mean weight was 75.1 kg (42.5–125), serum creatinine 1.12 mg/dL (0.33–4.41), serum albumin 3.5 g/dL (2.3–4.6) and haematocrit 32.6% (13.4–48.5). None of the patients received the registered drugs.

The one-compartment model showed a better OFV than the two-compartment model (-663,636 vs -654,430). However, the graphical analysis showed a better correlation between the CSC and those predicted, therefore the analysis of the covariates was continued with the two-compartment model.

The variables were evaluated in the two-compartment model and an influence of age and weight on clearance was observed, with statistically insignificant differences. No covariate showed an effect on the volume of distribution.

Conclusion and Relevance

• The two-compartment model with first order conditional estimation with interactions showed a better goodness of fit.

• The development of a pharmacokinetic model of cyclosporine assists clinicians to establish an effective and safe dosing regimen.

• Further studies are needed to better analyse the population pharmacokinetics of cyclosporine.

Conflict of Interest No conflict of interest.