Marije E. Otto [1,2] Laura G.J.M. Borghans [1], Joost van Mechelen [1,3], Gabriel E. Jacobs [1,4], J. G. Coen van Hasselt [2], Linda B.S. Aulin [1,2]
[1] Centre for Human Drug Research (CHDR), Leiden, The Netherlands [2] Leiden Academic Centre for Drug Research (LACDR), Leiden University, Leiden, The Netherlands [3] Leiden University Medical Centre (LUMC), Leiden, The Netherlands [4] Department of Psychiatry, Leiden University Medical Centre (LUMC), Leiden, The Netherlands
Objectives: The interest in esketamine as antidepressant drug remains high ever since it first got registered as an option for treatment resistant major depressive disorder[1]. Esketamine demonstrates rapid and sustained antidepressant effects lasting up to two weeks after a single dose. Esketamine administration via intravenous (IV) infusion has been studied the most, but is less feasible clinically. Intranasal and oral (PO) administration are being explored as alternative routes of administration[2]. Although PO adminstration comes with the advantage of being cost-effective, a great disadvantage is esketamine’s extensive first pass effect, leading to increased pharmacokinetic (PK) variability and a high metabolite to parent concentration ratio as compared to IV[3]. The currently available population PK (popPK) model for PO esketamine solution does not fully reach the therapeutic exposure range for antidepressant use and no between-occasion variability was quantified, thereby limiting its use for clinical trial simulations[4,5]. Therefore, we aimed to characterize the absorption PK and related variability of PO esketamine in healthy volunteers over a therapeutic dose range, by extension of a previously developed IV esketamine popPK model[6].
Methods: Data from a clinical trial investigating IV and PO administered esketamine solution (Ketanest®) in 17 healthy volunteers were used. Single doses of 0.4 mg/kg IV infusion over 40 min, 0.20 mg/kg PO and 0.45 mg/kg PO were administered in a crossover design. Total esketamine concentration levels were measured from venous blood plasma samples, taken at 11 timepoints up to 24h post dose.
A sequential modelling approach was used, where a previously reported model describing IV esketamine PK was fitted to the IV data to determine the empirical Bayes estimates (EBE’s)[6]. The model consisted of 3 distribution compartments, with inter-individual variability (IIV) on central clearance (CL), central distribution volume (Vd), and intercompartmental CL, and between-occasion variability (BOV) on CL. Individual distribution and elimination parameters were fixed to their EBE’s and different structures were tested to describe the absorption kinetics while bioavailability (F) was estimated, including 0-order, 1-order and combined absorption, and transit compartments to account for delay. IIV and BOV on absorption parameters were included in a stepwise manner. Model development steps were evaluated by drop in objective function value (OFV, p<0.01), relative standard error (RSE<50%), condition number, and goodness-of-fit (GOF) plots. Model development was done using NONMEM (V7.5), data transformation and visualization was done in R (V4.3.2).
Results: Esketamine absorption was best described with saturable absorption (k50%, 235 µg; kmax, 7205 µg/h with 90% IIV), with one transit compartment prior to the absorption compartment. In line with literature, a low F was estimated (13%) with high IIV (50%) and BOV (23%)[2]. For a typical individual, 92% of the maximum absorption rate was reached for the high PO dose versus 80% for the low PO dose. Inclusion of BOV on F increased RSE’s of transit rate and IIV estimates to >50%, but significantly improved the OFV and GOF, and was included in the model. Exploration of kmax and F EBE’s revealed a potential relationship, where kmax increases with increasing F. Since F and kmax are both related to the first pass metabolism in the gut and liver, differences in intrinsic activity of CYP2B6 and CYP3A4 enzymes might explain this finding[7]. GOF plots showed that observations were predicted accurately for the low PO dose, but were overpredicted for higher concentrations and underpredicted for later timepoints for the high PO dose, which may be linked to enzyme capacity, having more impact than expected for the high PO dose.
Conclusion: The popPK modelling approach identified high IIV for F and kmax and BOV for F for PO esketamine, and required saturable kinetics to describe the data. This model predicts later maximum concentrations with increasing doses, which may have implications for future studies and warrants further testing of high PO doses. Still, model refinement is necessary to correctly capture the kinetics of the high PO dose. As absorption, first-pass metabolism and systemic clearance are interlinked, unravelling the individual contribution of each process to the observed non-linear PK may require a more semi-mechanistic modelling approach.
References:
[1] Johnston, J. N., Kadriu, B., Kraus, C., Henter, I. D. & Zarate, C. A. Ketamine in neuropsychiatric disorders: an update. Neuropsychopharmacology 49, 23–40 (2024).
[2] Dutton, M., Can, A. T., Lagopoulos, J. & Hermens, D. F. Oral ketamine may offer a solution to the ketamine conundrum. Psychopharmacology (Berl). 240, 2483–2497 (2023).
[3] Zanos, P. et al. Ketamine and ketamine metabolite pharmacology: Insights into therapeutic mechanisms. Pharmacol. Rev. 70, 621–660 (2018).
[4] Smith-Apeldoorn, S. Y. et al. Oral esketamine for treatment-resistant depression: Rationale and design of a randomized controlled trial. BMC Psychiatry 19, (2019).
[5] Ashraf, M. W., Peltoniemi, M. A., Olkkola, K. T., Neuvonen, P. J. & Saari, T. I. Semimechanistic Population Pharmacokinetic Model to Predict the Drug–Drug Interaction Between S-ketamine and Ticlopidine in Healthy Human Volunteers. CPT Pharmacometrics Syst. Pharmacol. 7, 687–697 (2018).
[6] Otto, M., Borghans, L.G.J.M. , Van Mechelen, J., Jacobs, G., Van Hasselt, J.G.C.; Integrative pharmacokinetic model for esketamine and its metabolite esnorketamine, PAGE 31 (2023) Abstr 10594 [www.page-meeting.org/?abstract=10594]]
[7] Li, Y. et al. The CYP2B6*6 allele significantly alters the N-demethylation of ketamine enantiomers in vitro. Drug Metab. Dispos. 41, 1264–1272 (2013).
Reference: PAGE 32 (2024) Abstr 10777 [www.page-meeting.org/?abstract=10777]
Poster: Drug/Disease Modelling - Absorption & PBPK