Authors :
Presenting Author: Xintian Lyu, BS – University of Minnesota
Yuhan long, BS – Uniersity of Minnesota
Angela Birnbaum, PhD – University of Minnesota
Ilo Leppik, MD – University of Minnesota
Jeannine Conway, PharmD – University of Minnesota
Sílvia Illamola, PhD, PharmD – University of Minnesota
Rationale:
Levetiracetam (LEV) concentration was reported to be significantly higher in older adults than in younger adults with epilepsy. Age-related physiological changes, especially reduced renal function, can lead to decreased LEV clearance and result in higher drug concentrations, potentially increasing the risk of adverse events. This highlights the potential need for dose adjustments in older populations. This study utilized a population pharmacokinetic (popPK) modeling approach to characterize the pharmacokinetic profile of LEV in younger and older adults with epilepsy. The model uses covariate testing to account for interindividual variability (IIV) and supports dose simulations to inform future dose optimization strategies.
Methods:
LEV dosing, concentration, and covariate data (age, sex, race, weight, serum creatinine) were extracted from Fairview Health/University of Minnesota Physicians electronic health records between January 1, 2015 and June 13, 2023. Individuals were included if they were taking LEV, had dosing information, and had at least one concentration associated with an immediate-release oral formulation collected within 24 hours of the last dose. Concentrations measured during pregnancy, within one year postpartum, or below the lower limit of quantification (0.2 µg/mL) were excluded. Creatinine clearance was calculated using the Cockcroft–Gault equation. Additive, proportional, and combined residual error models were tested. Covariates evaluated included weight, sex, race, creatinine clearance, and age. PopPK modeling was performed using NONMEM with PsN (version 7.5.0).
Results:
A total of 909 LEV concentrations from 704 individuals (51.0% female) were included. Participants contributed a median of 1 and up to 13 concentrations each. Of all concentrations, 61.6% were from individuals aged < 65 years (mean age: 44.8), and 38.4% were from individuals aged ≥65 years (mean age: 77.6). The absorption rate constant (Ka) was fixed at 4.8 h-1, based on published literature. A one-compartment model with a combined residual error structure better described the concentration–time data. IIV was included on both clearance and volume of distribution. Creatinine clearance was identified as the most significant covariate, and it was retained on clearance in the model.
Conclusions:
Creatinine clearance on clearance was identified as the most significant covariate. Age influences LEV clearance through its association with renal function, demonstrating differences in LEV pharmacokinetics between younger and older adults with epilepsy. We recommend that serum creatinine be measured, at minimum, in older adults with epilepsy throughout LEV treatment. Future steps include evaluating additional covariates, such as interoccasion variability, BMI, and BSA, and simulating doses to provide individualized dosing strategies.
Funding: None