Pharmacokinetic modelling of orally administered cannabidiol and implications for medication control in horses

Abstract

Cannabidiol (CBD) products gain increasing popularity amongst animal owners and veterinarians as an alternative remedy for treatment of stress, inflammation or pain in horses. Whilst the use of cannabinoids is banned in equine sports, there is limited information available concerning CBD detection times in blood or urine. The aim of this study was to determine the pharmacokinetic properties of CBD following oral administration in the horse to assist doping control laboratories with interpreting CBD analytical results. Part 1: dose escalation study: Single oral administration of three escalating doses of CBD paste (0.2 mg/kg, n = 3 horses; 1 mg/kg, n = 3; 3 mg/kg, n = 5) with >7 days wash-out periods in between. Part 2: multiple dose study: oral administration of CBD paste (3 mg/kg, n = 6) twice daily for 15 days. Multiple blood and urine samples were collected daily throughout both studies. Following study part 2, blood and urine samples were collected for 2 weeks to observe the elimination phase. Concentrations of CBD, its metabolites and further cannabinoids were evaluated using gas-chromatography/tandem-mass-spectrometry. Pharmacokinetic parameters were assessed via two approaches: population pharmacokinetic analysis using a nonlinear mixed-effects model and non-compartmental analysis. AUC_{0-12 h} and C_max were tested for dose proportionality. During the elimination phase, the CBD steady-state urine to serum concentration ratio (Rss) was calculated. Oral CBD medication was well-tolerated in horses. Based on population pharmacokinetics, a three-compartment model with zero-order absorption most accurately described the pharmacokinetic properties of CBD. High volumes of distribution into peripheral compartments and high concentrations of 7-carboxy-CBD were observed in serum. Non-compartmental analysis identified a C_max of 12.17 ± 2.08 ng/mL after single administration of CBD (dose: 3 mg/kg). AUC_{0-12 h} showed dose proportionality, increase for C_max leveled off at higher doses. Following multiple doses, the CBD terminal half-life was 161.29 ± 43.65 h in serum. Rss was 4.45 ± 1.04. CBD is extensively metabolized and shows high volumes of tissue distribution with a resulting extended elimination phase. Further investigation of the potential calming and anti-inflammatory effects of CBD are required to determine cut-off values for medication control using the calculated Rss.

Keywords: CBD; Monolix; NLME model; PK; cannabinoids; doping; drug control; equine.

Figure 1

Mean ± standard deviation of serum and urine concentrations of cannabidiol (CBD) and the metabolites 7-hydroxy-cannabidiol (7-OH-CBD) and 7-carboxy-cannabidiol (7-COOH-CBD) after single oral administration of CBD paste in three different doses [0.2 mg/kg (A,B); 1 mg/kg (C,D); 3 mg/kg (E,F)].

Figure 2

Mean ± standard deviation of serum (A) and urine (B) concentrations of cannabidiol (CBD) and the metabolites 7-hydroxy-cannabidiol (7-OH-CBD) and 7-carboxy-cannabidiol (7-COOH-CBD) following multiple administrations of CBD paste (3 mg/kg po) twice daily over 2 weeks with subsequent sample collection.

Figure 3

Diagnostic plots extracted from the three-compartment model following population pharmacokinetic analysis. (A) Plot of observations vs. individual predictions. Blue dots indicate observations, red dots indicate censored data, black line—identity line; dotted black line represents the 90% prediction interval. Outliers proportion was 10.54%. (B) Scatterplot of individual weighted residuals (IWRES) vs. individual predictions. Blue dots indicate observations, red dots indicate censored data, spline is marked with a yellow line.

Figure 4

Diagnostic plot extracted from the three-compartment model following population pharmacokinetic analysis: visual predictive check for CBD concentrations in serum. Empirical data [10th, 50th (median) and 90th percentile] are marked by solid lines. Outlier dots are circled in red. Shaded areas mark the 90% confidence intervals for corrected prediction of the median (red) and the 10th and 90th percentile (blue).

Figure 5

Diagnostic plots extracted from the three-compartment model following population pharmacokinetic analysis: exemplary individual predictions for concentrations of cannabidiol (CBD) in serum after single oral administration of CBD paste in three different doses [(A): 0.2 mg/kg po; (B): 1 mg/kg po; (C): 3 mg/kg po], and (D): following multiple administrations of CBD paste (3 mg/kg po) twice daily over 2 weeks with subsequent sample collection. Green lines represent CBD administrations, blue dots are observed data points and black lines are individual fits.

Figure 6

Diagnostic plot extracted from the three-compartment model following population pharmacokinetic analysis: Correlation plots of the random effects (η_i). Correlation was applied when correlation coefficients were estimated to be high and met the threshold for inclusion (Pearson’s correlation test, p < 0.05). Linear regressions are presented as red lines.

Figure 7

Mean ± standard deviation of serum and urine concentrations of cannabidiol (CBD) during the elimination phase. Last CBD administration (dose: 3 mg/kg) to six horses at time point 336 h following multiple administrations of CBD paste (3 mg/kg po) twice daily over 2 weeks. Numbers present the urine/serum ratio between respective time points.