Effects of Genetic Polymorphism in CYP2D6, CYP2C19, and the Organic Cation Transporter OCT1 on Amitriptyline Pharmacokinetics in Healthy Volunteers and Depressive Disorder Patients

Abstract

The tricyclic antidepressant amitriptyline is frequently prescribed but its use is limited by its narrow therapeutic range and large variation in pharmacokinetics. Apart from interindividual differences in the activity of the metabolising enzymes cytochrome P450 (CYP) 2D6 and 2C19, genetic polymorphism of the hepatic influx transporter organic cation transporter 1 (OCT1) could be contributing to interindividual variation in pharmacokinetics. Here, the impact of OCT1 genetic variation on the pharmacokinetics of amitriptyline and its active metabolite nortriptyline was studied in vitro as well as in healthy volunteers and in depressive disorder patients. Amitriptyline and nortriptyline were found to inhibit OCT1 in recombinant cells with IC₅₀ values of 28.6 and 40.4 µM. Thirty other antidepressant and neuroleptic drugs were also found to be moderate to strong OCT1 inhibitors with IC₅₀ values in the micromolar range. However, in 35 healthy volunteers, preselected for their OCT1 genotypes, who received a single dose of 25 mg amitriptyline, no significant effects on amitriptyline and nortriptyline pharmacokinetics could be attributed to OCT1 genetic polymorphism. In contrast, the strong impact of the CYP2D6 genotype on amitriptyline and nortriptyline pharmacokinetics and of the CYP2C19 genotype on nortriptyline was confirmed. In addition, acylcarnitine derivatives were measured as endogenous biomarkers for OCT1 activity. The mean plasma concentrations of isobutyrylcarnitine and 2-methylbutyrylcarnitine were higher in participants with two active OCT1 alleles compared to those with zero OCT1 activity, further supporting their role as endogenous in vivo biomarkers for OCT1 activity. A moderate reduction in plasma isobutyrylcarnitine concentrations occurred at the time points at which amitriptyline plasma concentrations were the highest. In a second, independent study sample of 50 patients who underwent amitriptyline therapy of 75 mg twice daily, a significant trend of increasing amitriptyline plasma concentrations with decreasing OCT1 activity was observed (p = 0.018), while nortriptyline plasma concentrations were unaffected by the OCT1 genotype. Altogether, this comprehensive study showed that OCT1 activity does not appear to be a major factor determining amitriptyline and nortriptyline pharmacokinetics and that hepatic uptake occurs mainly through other mechanisms.

Keywords: CYP2C19; CYP2D6; OCT1; SLC22A1; amitriptyline; drug transport; nortriptyline; organic cation transporter 1.

FIGURE 1

Schematic illustration of the processes that determine the pharmacokinetics of AT and NT. P-glycoprotein (P-gp, MDR1, ABCB1), as part of the blood-brain-barrier, was shown in mice to transport AT and NT from the central nervous system into brain capillaries, thereby determining their concentrations at the synapse (Uhr, 2000; Grauer and Uhr, 2004; Uhr et al., 2007).

FIGURE 2

In vitro assessment of OCT1 inhibition by a range of different antidepressant and neuroleptic drugs. Shown is the mean negative logarithm of the IC₅₀ values of 3–4 independent experiments, the error bars indicate the SEM.

FIGURE 3

Individual plasma concentrations of (A) AT and (B) NT over time for all healthy volunteers (black curves), indicating the large interindividual variation for these tricyclic antidepressants. The mean (±SEM) concentrations for carriers of two (green), one (yellow), or zero (red) active OCT1 alleles are superimposed (OCT1-dependent differences in AUC were not statistically significant). The single participant with the highest NT concentrations had wild-type OCT1 genotype, reduced activity CYP2D6 genotype, and a very high activity CYP2C19 genotype.

FIGURE 4

Mean plasma concentrations of AT (left) and NT (right) in healthy volunteers over time, stratified by the number of active alleles for CYP2D6 (top) and CYP2C19 (bottom).

FIGURE 5

Intensity of fatigue after AT administration reported by the participants using a visual analogue scale. Shown is the time course for each participant (black curves) and the mean superimposed (red curve). The intensity of fatigue was not dependent on OCT1, CYP2D6, or CYP2C19 genotypes (p > 0.05, Jonckheere-Terpstra test).

FIGURE 6

Mean plasma concentrations per dose unit of (A) AT and (B) NT, stratified by OCT1 genotype, in 50 depressive disorder patients who underwent AT therapy.

FIGURE 7

(A) Mean ± SEM of IBC (dark coloured circles) and AT (light coloured squares) plasma concentrations over time, stratified by OCT1 genotype (green data points represent two active OCT1 alleles and red data points represent zero active OCT1 alleles; carriers of OCT1*2 were not included). (B) Hysteresis plot showing the mean plasma concentrations of AT and IBC in 13 healthy volunteers with two active OCT1 alleles. (C) Mean ± SEM of 2-methylbutyrylcarnitine and (D) propionylcarnitine plasma concentrations over time, stratified by OCT1 genotype.

FIGURE 8

Different starting dosage adjustment recommendations from the literature and based on the results of this study for (A) CYP2D6 and (B) CYP2C19. These were taken from the Clinical Pharmacogenetics Implementation Consortium (CPIC^®) guideline (Hicks et al., 2017), the Dutch Pharmacogenetics Working Group (DPWG) guideline (August 2019 update), or calculated based on the formulas described by Stingl et al. (2013) and by using the AUCs determined in the respective studies. In accordance with the CPIC^® and DPWG final consensus on CYP2D6 genotype to phenotype (Caudle et al., 2020), a CYP2D6 activity score of 0 was classified in this study as poor, of 0.5 and 1 as intermediate, of 1.5 and 2.0 as normal/extensive, and of >2.5 as ultra-rapid metaboliser phenotypes. For CYP2C19, an activity score of 0 was classified in this study as poor, of 1 as intermediate, of 1.5 and 2 as normal/extensive, and of >2 as ultra-rapid metaboliser phenotypes. The starting dosage adjustment recommendations are also listed in Supplementary Table S4. As apparent, there is a high consistency between different recommendations and the measurements from this study, particularly with regard to the CYP2D6 genotype.