Integrative pharmacokinetic model for esketamine and its metabolite esnorketamine
Marije E. Otto [1,2], Laura G.J.M. Borghans [1], Joost van Mechelen [1], Gabriel Jacobs [1,3], J. G. Coen van Hasselt [2]
[1] Centre for Human Drug Research (CHDR), Leiden, The Netherlands [2] Leiden Academic Centre for Drug Research (LACDR), Leiden University, Leiden, The Netherlands [3] Department of Psychiatry, Leiden University Medical Centre (LUMC), Leiden, The Netherlands
Objectives
One might say that the current resurgence of the drug development pipeline for new antidepressant treatments was initiated by esketamine, being a FDA and EMA registered treatment for treatment-resistant major depressive disorder since 2019 [1,2]. In contrast to classical antidepressants effects are rapid and already apparent after a single dose. However, the specific attribution of these clinical effects to esketamine and its metabolites, such as esnorketamine and eshydroxynorketamine, is topic of debate and requires further research, especially as administration route highly impacts the pharmacokinetic (PK) profile of both parent and metabolites due to an extensive first pass effect. Furthermore, different pharmacokinetic-pharmacodynamic (PKPD) relationships might also exist between parent and metabolites for the acute versus long term effects. With multiple analogues and adjusted formulations currently being developed, further investigation of these factors using PK(PD) modelling will allow for improved dose selection and frequency of future clinical trials. Therefore, the objective of this study was to characterize the PK of esketamine and its metabolites in healthy volunteers.
Methods
Study data
Pharmacokinetic data from three clinical studies investigating intravenously administered esketamine in healthy volunteers was available [3,4]. All studies measured esketamine (KET) and esnorketamine (NOR) blood levels. The administered doses ranged from 0.15 to 0.83 mg/kg in total, administered over 30 to 120 minutes with and without a bolus loading dose. Both single and multiple treatment period data was available. In total, 815 samples from 65 subjects were collected up to 24h after start of dosing.
Model development
Population PK model development was done in a sequential fashion. First a PK model was developed for the parent drug KET. In a second step, metabolite models for NOR and HNOR were developed while the parent drug KET was kept fixed. Multiple structural models were explored, including transit compartments to account for delay in metabolite formation. Inter-individual and between-occasion variability (IIV, BOV) were included in a stepwise manner. Distribution volume (Vd) and clearance (CL) parameters were allometrically scaled a priori. NOR central Vd was fixed to KET Vd due to parameter identifiability issues. The M3 method was used to include data below the limit of quantification [5] and the L2 method was used to include correlation between KET and NOR residual unexplained variability [6]. Decisions regarding model development were supported by drop in objective function value (dOFV ≤ -6.64, p = 0.001), relative standard error (RSE<50%), condition number (<1000), goodness-of-fit (GOF) plots and a prediction corrected visual predictive check (pcVPC).
Software
Model development was performed using NONMEM (V7.5) while data transformation and visualization was done in R (V4.0.3)[6,7].
Results
KET and NOR distribution was described with 3 and 2 compartment models, respectively. Use of one transit compartment to describe metabolite formation improved the model significantly (dOFV=-7.26). IIV was included on central Vd, CL and intercompartmental CL and BOV was included for CL for both analytes. Goodness-of-fit plots and pcVPC plots showed that the model was able to accurately predict the observed data. However, exploration of Empirical Bayes Estimates and GOF plots indicated a study specific trend in CL.
Conclusion
A population PK model describing the PK of esketamine and its metabolites after IV administration was developed. As one of the clinical studies also provides eshydroxynorketamine and oral administration data, an extension of the model including this analyte and absorption kinetics is planned for the near future. This model will provide the basis for further PKPD modeling to investigate and quantify the pharmacological of effects to esketamine and its metabolites, and can support drug development and dose optimization of esketamine.
References:
[1] Spravato (esketamine). European Medicines Agency https://www.ema.europa.eu/en/medicines/human/EPAR/spravato#authorisation-details-section (2019).
[2] FDA approves new nasal spray medication for treatment-resistant depression; available only at a certified doctor’s office or clinic. FDA News Release https://www.fda.gov/news-events/press-announcements/fda-approves-new-nasal-spray-medication-treatment-resistant-depression-available-only-certified (2019).
[3] Okkerse, P. et al. The use of a battery of pain models to detect analgesic properties of compounds: a two-part four-way crossover study. Br. J. Clin. Pharmacol. 83, 976–990 (2017).
[4] Kleinloog, D., Stevens, J., Heuberger, J., Spinhoven, P. & van Gerven, J. The influence of personality on the sensitivity to subjective effects of δ9-tetrahydrocannabinol. Psychiatry Res. 220, 945–953 (2014).
[5] Beal, S. L. Ways to fit a PK model with some data below the quantification limit. J. Pharmacokinet. Pharmacodyn. 28, 481–504 (2001).
[6] Beal, S. L., Sheiner, L. B. & Boeckman, A. J. NONMEM 7.5.0 User Guides. (1989-2020). ICON Dev Solut Hanover, MD.
[7] R Core Team. R: A language and environment for statistical computing. R Found. Stat. Comput. Vienna, Austria (2020).