A PBPK model of 4β-hydroxycholesterol, an endogenous biomarker for hepatic CYP3A activity.
Aneesh Karkhanis (1), Matthew Harwood (1), Frederic Bois (1), Sibylle Neuhoff (1)
(1) Certara UK Limited, Simcyp Division, Level 2-Acero, 1 Concourse Way, Sheffield, S1 2BJ, UK
Introduction: New drug candidates are tested for CYP3A enzyme induction in vivo using specific probe substrates. However, due to the cost and invasive nature of such clinical trials, there is an increased desire to explore biomarkers that reflect CYP3A induction in vivo. Several studies have shown that plasma 4β-hydroxycholesterol (4β-OHC) levels increase in the presence of probe CYP3A inducers like rifampicin and carbamazepine [1, 2]. 4β-OHC is a CYP3A-specific metabolite of cholesterol that has received increased attention as an endogenous biomarker for hepatic CYP3A activity. It is primarily generated in the liver by CYP3A and is further metabolised by CYP7A1 and CYP27A1 [3].
Objectives: To develop a physiologically based pharmacokinetic (PBPK) model of cholesterol and 4β-OHC able to reproduce the clinically observed changes in plasma 4β-OHC levels due to hepatic CYP3A induction.
Methods: The extent of induction is dependent on the baseline 4β-OHC levels. Thus, the PubMed database was searched for papers to define such 4β-OHC baseline levels. Studies reporting values for healthy, adult individuals from different ethnic populations were collated. Studies in diseased populations or where only cholesterol/4β-OHC ratio was reported were excluded. Next, a minimal PBPK model was developed for cholesterol and 4β-OHC with constant infusion dosing in Simcyp PBPK Simulator (Version 23, Certara Ltd. UK, Sheffield). Relevant physicochemical properties and the blood-to-plasma ratio for cholesterol were obtained from the literature and it was assumed to be same for 4β-OHC. The fraction unbound for both compounds was predicted using in-built tools, while the plasma binding component was modified to reflect the binding to lipoprotein complexes [4]. A user-input volume of distribution value of 0.1 L/kg and 0.05 L/kg (%CV=15) were used for cholesterol and 4β-OHC, respectively. The liver:plasma partition coefficients for cholesterol and 4β-OHC were calculated from preclinical study data. The hepatic abundance of CYP27A1, derived from proteomics data was incorporated into the model to account for alternate metabolic pathways. The fraction of cholesterol metabolised in the liver by CYP3A (fm,CYP3A) and CYP27A1 (fm,(CYP27A1)) for cholesterol and 4β-OHC were optimised based on plasma cholesterol and 4β-OHC levels. The intrinsic clearance values for CYP450s were derived from published plasma half-lives of cholesterol and 4β-OHC and the respective fm values.
Results: Our meta-analysis revealed that the mean ± SD steady-state plasma concentration of 4β-OHC is 29.85 ± 14.87 ng/mL in healthy individuals (38 studies, 4466 individuals). The PBPK model recovered baseline plasma 4β-OHC levels in Caucasian (3 studies), Japanese (2 studies), Korean (2 studies), North American White (2 studies), Asian (1 study), African American (2 studies), and Latino (1 study) populations. The model also captured the higher baseline 4β-OHC levels in females compared to males, indicative of sex-specific differences in CYP3A abundances in the population. Similarly, the model recovered the higher baseline 4β-OHC levels in CYP3A5 extensive metabolisers compared to poor metabolisers in the North American Asian population. More importantly, the model recapitulated the increased 4β-OHC levels after multiple dose rifampicin treatment in four independent studies with an average fold error and absolute average fold error of 0.98 and 1.02 respectively.
Conclusion: The PBPK model recovered the baseline plasma 4β-OHC levels in different ethnicities; sex-specific and CYP3A5 phenotypic differences in baseline levels and increased levels in rifampicin-mediated hepatic CYP3A induction.
References:
[1] S. Kasichayanula, D.W. Boulton, W.L. Luo, A.D. Rodrigues, Z. Yang, A. Goodenough, M. Lee, M. Jemal, F. LaCreta, Validation of 4beta-hydroxycholesterol and evaluation of other endogenous biomarkers for the assessment of CYP3A activity in healthy subjects, Br J Clin Pharmacol, 78 (2014) 1122-1134.
[2] K. Bodin, L. Bretillon, Y. Aden, L. Bertilsson, U. Broome, C. Einarsson, U. Diczfalusy, Antiepileptic drugs increase plasma levels of 4beta-hydroxycholesterol in humans: evidence for involvement of cytochrome p450 3A4, J Biol Chem, 276 (2001) 38685-38689.
[3] K. Bodin, U. Andersson, E. Rystedt, E. Ellis, M. Norlin, I. Pikuleva, G. Eggertsen, I. Bjorkhem, U. Diczfalusy, Metabolism of 4 beta -hydroxycholesterol in humans, J Biol Chem, 277 (2002) 31534-31540.
[4] S.O. Olofsson, G. Bjursell, K. Bostrom, P. Carlsson, J. Elovson, A.A. Protter, M.A. Reuben, G. Bondjers, Apolipoprotein B: structure, biosynthesis and role in the lipoprotein assembly process, Atherosclerosis, 68 (1987) 1-17.