Quantification and Prediction of Human Fetal (-)-Δ9-tetrahydrocannabinol (THC)/11-OH-THC Exposure to Inform Neurodevelopmental Toxicity of Cannabis

Abstract

Prenatal cannabis use is associated with various detrimental outcomes in the offspring, the foremost being neurodevelopmental toxicity1-5. Δ9-tetrahydrocannabinol (THC) is the most abundant psychoactive compound in cannabis and is thought to be the cause of this toxicity. THC’s primary in vivo psychoactive metabolite, 11-OH-THC, may also contribute to this toxicity6. It is unethical to conduct a randomized clinical trial to determine if these cannabinoids are the causative factors in producing neurodevelopmental toxicity from prenatal cannabis use. An alternative approach is to conduct preclinical fetal cannabis toxicity studies in vivo in animals or in vitro at THC/11-OH-THC concentrations that mimic human fetal THC/11-OH-THC exposure. To determine the latter, we quantified the fetal circulation and tissue cannabinoid exposure across gestation. Since these determinations provide cannabinoid exposure at only a given time, we also developed a maternal-fetal physiologically based pharmacokinetic (m-f-PBPK) model to predict the time-dependent fetal THC/11-OH-THC exposure after chronic inhalation or oral THC consumption at various gestational ages. To accomplish the above goals, we first determined all the potential mechanisms of clearance and distribution that can affect THC/11-OH-THC fetal exposure. In Chapter 27, we incubated 500 nM THC and 50 nM 11-OH-THC in adult human intestine, lung, placenta, and in fetal liver microsomes with selective inhibitors for CYP (1A, 2C19, 2C9, 2D6, 3A), UGT (1A9, 2B7), and/or FMO enzymes to identify the drug metabolizing enzymes responsible for depletion of the cannabinoids. While there was no significant depletion of either cannabinoid in the human lung or placenta, in the intestine THC was significantly metabolized by CYP2C9 (89%) and CYP3A (11%) while 11-OH-THC was significantly metabolized by CYP3A (51%) and UGT2B7 (25%). In the fetal liver, both compounds were nearly completely metabolized by CYP3A7. The presence of intestinal and fetal liver metabolism, respectively, will impact bioavailability after oral THC administration and fetal exposure, relative to the maternal exposure. Next, in Chapter 38, we determined the mechanism and quantified the extent of transplacental transfer of THC and its metabolites. To do so, we perfused THC alone (5 μM) or in combination (100 – 250 nM) with its metabolites (100 nM or 250 nM 11-OH-THC, 100 nM COOH-THC) in an ex vivo dual cotyledon, dual perfusion, term human placenta model. While 11-OH-THC and COOH-THC appeared to passively diffuse across the placenta, 5 μM THC had a significantly lower maternal-fetal vs fetal-maternal clearance, suggesting that THC is actively transported, in the fetal-maternal direction, at the placental barrier. When a P-gp/BCRP inhibitor (4 μM valspodar) was added to the perfusion, the difference in clearance remained, suggesting that THC was transported by an unknown apical efflux or basal influx transporter(s) in the syncytiotrophoblast membrane. The estimated human fetal microsomal metabolism (Chapter 2) and the passive diffusion/active transport kinetics of THC (Chapter 3), which both reduce fetal cannabinoid exposure, were inputted in our m-f-PBPK model to predict fetal plasma and brain cannabinoid concentrations in trimester 1 (T1), 2 (T2), and 3 (T3) (Chapter 4). These predictions aligned well with the corresponding values observed in vivo, i.e. our model is considered verified. For example, the predicted umbilical venous plasma (UVP)/maternal plasma (MP) at gestational week (GW) 38 and fetal brain/MP at GW15, 0.26 and 0.56 respectively, were similar to the observed values in vivo, 0.33 (38 %CV) and 0.39 (63 %CV) respectively. Then, we used our m-f-PBPK model to predict the steady-state total and unbound fetal brain and plasma THC/11-OH-THC concentrations across gestational age. The maximum (Cmax,ss) and average (Css,avg) steady concentrations in the UVP and fetal brain of both THC and 11-OH-THC were highest at the earliest GW tested, 15 weeks. For example, at GW15, after a typical daily dose of THC consumed by inhalation (100 mg), the predicted fetal brain Css,avg of THC and 11-OH-THC was 37 and 70 nM, respectively. After a typical daily oral consumption of 10 mg THC, the predicted Css,avg of THC and 11-OH-THC in the fetal brain was lower, 0.73 and 8.9 nM, respectively. The corresponding unbound fetal brain Css,avg,u was predicted to be even lower due to extensive THC binding in the brain. Our verified m-f-PBPK model can be used to predict these fetal brain concentrations for any dose, route or frequency of consumption. We propose that future fetal toxicity studies (in vitro or in vivo in animals) should be conducted to replicate the predicted unbound fetal UVP and fetal brain THC/11-OH-THC concentrations at the typical doses consumed by pregnant people. Such preclinical studies will provide a better understanding of the potential of THC/11-OH-THC to produce neurotoxicity in the offspring as a result of prenatal cannabis use.