Marloes B. van Leuken (1,2), Marije E. Otto (1,2), Katelijne V. van der Heijden (1,3), Gabriël E. Jacobs (1,4), J.G. Coen van Hasselt (2), A. Inamdar (5), Linda B.S. Aulin (1)
(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, (5) Cybin Inc., Toronto, Canada
Introduction: Psychedelic agents such as N,N-dimethyltryptamine (DMT) are currently being studied for the treatment of various mental disorders [1][2]. To guide further development as a potential therapeutic agent, characterization of the pharmacokinetics (PK) of DMT is essential. To circumvent extensive first-pass metabolism by monoamine oxidase A (MAO-A), DMT can be administered through intravenous (IV) infusion. After IV administration, the systemic clearance of DMT is high, also due to metabolism by MAO-A. Previously, a population PK (popPK) model was developed based on healthy subjects who received an IV bolus dose of DMT administered over 30 seconds [3]. However, clinical application also focuses on continuous IV infusion of DMT. It is currently unclear whether continuous infusion regimens may show differences in PK, such as non-linear kinetics. Furthermore, the effect of smoking may impact MAO-A activity and hence, could influence the PK of DMT [4]. Therefore, the aim of this research is to quantitatively characterize DMT PK following IV infusion in healthy smokers and non-smokers through development of a popPK model.
Methods: Study design: PK data were available from a clinical study investigating different infusion regimens of DMT hemifumurate in 24 healthy smokers (5-10 cigarettes daily) and 12 non-smokers. In the smoker cohorts, 0.12 mg/kg, 18.4 mg, 36.4 mg, or 72.8 mg was administered (N=6 per cohort) as a continuous infusion over 90 minutes in a single-ascending dose manner. Total DMT plasma concentrations were measured at 13 timepoints over the course of 180 minutes. In non-smokers, an IV bolus loading dose (18.2 mg) was administered over 5 minutes followed by continuous infusion over 55 minutes (44.5 or 71.2 mg) in a crossover ascending-dose design. Here, concentrations were measured at 14 timepoints over 150 minutes.
Model development: One-, two-, and three-compartmental models with linear and non-linear elimination and distribution were investigated. Inter-individual variability (IIV) on structural model parameters was included in a stepwise manner. Proportional and/or additive error models for residual unexplained variability were evaluated. The potential covariate smoking was explored by visual inspection of the empirical bayes estimates (EBEs). The model was evaluated based on the drop in objective function value, for nested models, (p<0.01), parameters’ relative standard error (RSE; <50%), condition number (<1000), goodness-of-fit (GOF) plots, and visual predictive checks. The popPK model was developed using NONMEM (V7.5.0) and PsN (V5.2.6) [5][6]. Data transformation and visualization was done using R (V4.0.3) [7].
Results: The PK of DMT was best described with a two-compartmental model with linear elimination and a proportional residual error of 0.15. The GOF plots indicated a good fit with exception for underpredicting the higher concentrations observed in smokers towards the end of the infusion. Including non-linear elimination or distribution did not improve the bias or overall model fit.
The population estimate for clearance (CL) was 1137.5 L/h (RSE 8.7%) and the estimate for distribution volume (Vd) was 222.7 L (RSE 11.6%), displaying rapid elimination and extensive distribution, respectively. The peripheral compartment had a volume of 38.4 L (RSE 11.6%) and intercompartmental clearance of 63.3 L/h (RSE 12.8%). IIV was included on CL and Vd, with coefficients of variation of 51.5% (RSE 25.2%) and 72.1% (RSE 23.8%), respectively. The preliminary covariate exploration showed that EBEs of smokers were lower for CL and higher for Vd as compared to non-smokers, indicating that smoking could potentially explain part of the IIV.
Conclusions: The developed popPK model describes the PK data adequately. Additional IIV on Vd could be quantified as compared to the literature model [3]. The cause of the unexpected bias in model predictions towards the end of the infusion remains unclear. The large differences between smokers and non-smokers could possibly be due to the reduced MAO-A activity in smokers or the distinct infusion regimen, i.e. continuous infusion preceded by a bolus loading dose or not [4]. This could also explain the substantial differences in population estimates between the developed and literature model [3]. Further research should investigate MAO-A activity as a potential covariate and explore the possible influence of infusion regimens on the disposition of DMT.
References:
[1] K. A. A. Andersen, R. Carhart‐Harris, D. J. Nutt, and D. Erritzoe, “Therapeutic effects of classic serotonergic psychedelics: A systematic review of modern‐era clinical studies,” Acta Psychiatr. Scand., vol. 143, no. 2, pp. 101–118, Feb. 2021, doi: 10.1111/acps.13249.
[2] H. Lowe et al., “Psychedelics: Alternative and Potential Therapeutic Options for Treating Mood and Anxiety Disorders,” Molecules, vol. 27, no. 8, p. 2520, Apr. 2022, doi: 10.3390/molecules27082520.
[3] E. Eckernäs, C. Timmermann, R. Carhart‐Harris, D. Röshammar, and M. Ashton, “Population pharmacokinetic/pharmacodynamic modeling of the psychedelic experience induced by N,N‐dimethyltryptamine – Implications for dose considerations,” Clin. Transl. Sci., vol. 15, no. 12, pp. 2928–2937, 2022, doi: 10.1111/cts.13410.
[4] J. S. Fowler et al., “Brain monoamine oxidase A inhibition in cigarette smokers,” Proc. Natl. Acad. Sci., vol. 93, no. 24, pp. 14065–14069, Nov. 1996, doi: 10.1073/pnas.93.24.14065.
[5] ICON Early Phase, “NONMEM®.” Accessed: Oct. 05, 2023. [Online]. Available: https://www.iconplc.com/solutions/technologies/nonmem/
[6] L. Lindbom, J. Ribbing, and E. N. Jonsson, “Perl-speaks-NONMEM (PsN) – A Perl module for NONMEM related programming,” Comput. Methods Programs Biomed., vol. 75, no. 2, pp. 85–94, 2004, doi: 10.1016/j.cmpb.2003.11.003.
[7] R Core Team, “R: A language and environment for statistical computing.” Accessed: Oct. 05, 2023. [Online]. Available: https://www.r-project.org/
Reference: PAGE 32 (2024) Abstr 10779 [www.page-meeting.org/?abstract=10779]
Poster: Drug/Disease Modelling - CNS