Cortisol disposition and cortisol-mediated drug interactions, including enterohepatic recycling and CYP3A induction in pregnant and premenopausal healthy women

Abstract

Prenatal cannabis exposure is associated with adverse fetal outcomes including impairedneurodevelopment and reduced birth weight. THC and its active metabolite 11-OH-THC are cleared primarily by hepatic CYP2C9, with a fraction metabolized (fm) of approximately 0.81, and the magnitude of fetal THC exposure depends directly on maternal hepatic clearance pathways. Controlled pharmacokinetic studies in pregnant individuals are challenging due to ethical and legal limitations, making mechanistic inference from endogenous biomarkers and carefully designed controlled hormone treatments a necessary approach. Cortisol, the primary endogenous glucocorticoid, increases 2- to 3-fold above non-pregnant plasmaconcentrations by the third trimester and has been proposed as a key inducer of hepatic CYP2C9 and CYP3A4 through glucocorticoid receptor activation. Cortisol also regulates the expression of hepatic and renal transporters through the glucocorticoid receptor–HNF4α transcriptional axis, with potential consequences for the elimination of glucuronide conjugates and organic cations. Critically, it is the unbound, pharmacologically active fraction of cortisol that drives receptor occupancy and downstream transcriptional effects. Thus accurate quantitation of unbound cortisol concentrations is essential to any mechanistic assessment of cortisol-driven DMET induction. The overarching hypothesis of this dissertation was that pregnancy-associated increases in cortisol exposure alter THC disposition by modulating DMET activity. To test this question has direct translational relevance: if cortisol drives the pregnancy-associated increases in CYP2C9 and CYP3A4 that are documented in vivo, then cortisol concentrations could serve as a mechanistic predictor of altered drug clearance across pregnancy. To test this hypothesis, Chapters 2–4 employed a randomized crossover clinical study in 13 healthypremenopausal women who received 7 days of oral hydrocortisone to recapitulate third trimester cortisol concentrations, achieving a 2.0-fold increase in total cortisol AUC and a 5.4-fold increase in unbound cortisol Cmax from 14 to 70 nM. This controlled design allowed isolation of the individual contribution of cortisol to DMET regulation, free from the complex hormonal milieu of pregnancy. Chapter 3 developed a plasma protein binding model (PPBM) using in vivo data from this clinical study to simultaneously characterize competitive binding of cortisol and cortisone to CBG and albumin, estimating in vivo dissociation constants of 13 nM for cortisol:CBG and 169 nM for cortisone:CBG. These were 3- to 10-fold lower than prior in vitro estimates, demonstrating that the native plasma environment substantially enhances CBG binding affinity and that reliance on in vitro estimates systematically overestimates the biologically active unbound fraction of cortisol. Despite the large increase in cortisol exposure, THC and 11-OH-THC exposure were completely unchanged in Chapter 2, and two orthogonal endogenous biomarker systems in Chapter 4, the 6β-hydroxylation of cortisol and cortisone and the DCA hydroxylation pathway, concordantly showed no meaningful induction of hepatic CYP3A activity. Instead, both chapters revealed consistent evidence of cortisol-mediated transporter induction. In Chapter 4, a selective 41% decrease in taurine-conjugated OH-DCA pointed to induction of biliary efflux transporters downstream of CYP3A formation, while a 5.2-fold increase in cortisol renal clearance far exceeding the change attributable to filtration alone indicated coordinated induction of renal OCT2 and MATE1 transport. In Chapter 2, cortisol treatment was further associated with a significant decrease in 11-COOH-THC glucuronide exposure and prolongation of the apparent half-life of 11-COOH-THC, consistent with induction of hepatic OCT1- mediated uptake and renal OCT2-mediated secretion of the cannabinoid glucuronide metabolite and increased enterohepatic recycling. Chapter 5 then extended these findings to 47 pregnant women during late pregnancy and postpartum, examining whether the mechanisms identified under controlled conditions translate to the full physiological complexity of pregnancy, where total cortisol increases 2.7-fold yet unbound cortisol increases only 2.0-fold owing to pregnancy-specific tightening of CBG binding affinity. Mirroring the findings from the controlled hydrocortisone study, renal clearance of cortisol and 6β- hydroxycortisol each increased during late pregnancy with filtration clearance unchanged, indicating induction of renal tubular secretion transporters consistent with the OCT2 and OAT3 induction observed under controlled conditions. Together, the chapters build from a deconvolution of the role of cortisol under controlled conditions to a quantitative characterization of enzyme and transporter activity in vivo during pregnancy.