Parsshava Mehta (1), Leyanis Rodriguez (1), Vijay K Siripuram (1), Paula Muniz (2), Monica Rodriguez (2 ), Emili González-Pérez (3), Marta Forcadell (3), Valvanera Vozmediano (1)
1) Department of Pharmaceutics, Centre for Pharmacometrics and Systems Pharmacology, University of Florida, Gainesville, Florida, USA., (2) Model Informed Development (MID), CTI Clinical Trial & Consulting, Bilbao, Spain. , (3) Clinical Trial Development, R&D Unit, Neuraxpharm, Barcelona, Spain.
Introduction: Adequate delivery of drugs at the central nervous system (CNS) level is a prerequisite for drugs in epilepsy. However, the measure of concentrations of antiepileptic drugs on target sites in humans is highly restricted and faces the challenge of the ethical limitations related to the invasive nature of sampling extracellular fluid (ECF) concentrations. A unique strategy to characterize the rate of entry through the blood-brain barrier (BBB) is provided by physiologically based pharmacokinetic modeling (PBPK) after integration of preclinical data.
Objectives: To characterize levetiracetam permeability across the BBB using PBPK modeling and to evaluate drug disposition at the target site and its pharmacodynamic response.
Methods: The anti-epileptic medication levetiracetam was studied using a PBPK model in rats, after integration of a CNS four-compartment PBPK model [1] in R® [2], which included the intracellular fluid (ICF), ECF, cranial cerebro-spinal fluid (CSF), and spinal CSF. The model was initially developed for the 40 mg/kg dosing obtained from Tong et al. [3], which provides the plasma and ECF concentrations in rats. The BBB permeability was optimized using a naive-pooled technique in PUMAS [4]. The model was externally verified for the 80 mg/kg dosing from Tong et al. [3] (plasma/ECF) and Doheny et al. [5] (plasma/CSF). The permeability surface area product (PSbbb) in rats was then scaled to account for human BBB passage [6]. The full human PBPK model predictions were compared with 1) observed plasma concentrations for levetiracetam immediate release (IR) [7], and 2) ratios plasma:CSF and CSF:ECF [8]. Total brain and CSF concentrations were then linked to a pharmacodynamic model using SVA2 receptor occupancy (RO) data from the literature [9]. The full PBPK/RO model was then used to simulate RO fluctuations and the required RO for effectiveness at steady state.
Results: Optimization of PSbbb product for the BBB was 8.7 times lower than the original PSbbb obtained using Caco-2 cells [10] (PSbbb_optim = 0.594 L/hr vs PSbbb_initial = 5.184 L/hr). A comparison for the ratio of observed (obs) vs predicted (pred) for 40 mg/kg & 80 mg/kg for AUC_serum = 0.89 & 0.96, Cmax_serum = 0.94 & 1.00, AUC_ECF = 1.21 & 1.20 and Cmax_ECF = 1.23 & 1.13, respectively, were within the 1.25 acceptance criteria. In humans, a comparison of ratio of obs vs pred for the ratios of Caverage: CSF/Plasma = 0.80 vs 0.72 and for Caverage: CSF/ECF = 3.98 vs 4.02 increased confidence on the model’s predictions for CSF and ECF [8]. A simulation to steady state (5 days) showed that the fluctuations in the plasma concentrations are minimized at the level of the brain, due to delayed penetration of levetiracetam. The model prediction showed that RO fluctuations at steady state were between 94-65% during the entire dosing interval.
Conclusion: Using this innovative PBPK model, we were able to predict levetiracetam BBB permeability using rat microdialysis experiments. Additionally, using the PBPK/RO approach, we were able to correlate the concentrations at the site of action to RO and calculate the SV2A RO for levetiracetam IR formulation. This approach may also help to determine whether the minimal RO required for clinical effectiveness is reached for different posologies by better understanding the relationship between exposure at the target site and pharmacodynamics of the drug.
References:
[1] Verscheijden LFM et al., PLoS Computational Biology, 2019;15(6):1–19.
[2] R Core Team (2020). R: A language and environment for statistical computing. https://www.R-project.org/.
[3] Tong X et al., British Journal of Pharmacology. 2001;133(6):867–74.
[4] Rackauckas C et al., bioRxiv 2020.11.28.402297.
[5] Doheny HC et al., Epilepsy Research. 1999;34(2–3):161–8.
[6] Simcyp. 17.0.90.0 ed. available from: https://www.certara.com/.
[7] https://www.accessdata.fda.gov/drugsatfda_docs/nda/2008/022285s000_ClinPharmR.pdf
[8] Rambeck B et al., Epilepsia. 2006;47(4):681–94.
[9] Finnema SJ et al., Epilepsia. 2019;60:958–967.
[10] Nicolas JM et al., Epilepsia, 57(2):201–209, 2016.
Reference: PAGE 30 (2022) Abstr 10180 [www.page-meeting.org/?abstract=10180]
Poster: Drug/Disease Modelling - CNS