Increased use of complex cellular systems to study hepatic drug disposition in vitro has been associated with increased complexity of models applied for data analysis. Despite these advances, a tendency to under-predict transporter-mediated hepatic clearance using in vitro data is generally reported. The aim of the current project was to investigate systematically the reasons for this under-prediction. Focus of the work was on differences in quantitative expression between Organic anion-transporting polypeptide (OATP) transporters in the in vitro cellular systems and human liver tissue and their impact on physiologically-based pharmacokinetic-in vitro-in vivo extrapolation (IVIVE) of transporter kinetic data. In addition, various modelling techniques to estimate kinetic parameters were explored, using existing and newly generated data in rat and human plated hepatocytes. Finally, to develop a novel experimental method for investigating uptake in a reduced experimental format, validate the method in rat hepatocytes and apply it to human and investigate the impact on IVIVE of transporter kinetic data. Nine OATP substrates were selected, namely rosuvastatin, valsartan, repaglinide, olmesartan acid, candesartan acid, fexofenadine, cerivastatin, glibenclamide and rosiglitazone. Meta-analysis of reported quantitative protein levels of OATP transporters in different human hepatocyte formats (suspended and sandwich-cultured hepatocytes) and liver tissue was performed. The analysis showed maximal 2.5-fold difference between OATP expression in hepatocytes and liver tissue (data from 86 individuals) which could not solely rationalize under-prediction of active uptake clearance reported so far. Weak correlation between OATP1B1 expression and age was noted, but no direct relationship between sex and OATP expression levels was observed. A new modelling strategy for fitting sparse rosuvastatin in vitro uptake data from human hepatocytes was explored. A mechanistic two-compartment model was set up NONMEM 7.3 allowing the simultaneous estimation of rosuvastatin kinetic parameters describing transported-mediated uptake (CLactive), bidirectional passive diffusion (CLpassive) and intracellular binding (fu,cell) from uptake experiments carried out over 2 to 60 min and over a range of concentrations (0.1 to 300 µM). Use of a population-based approach resulted in structural and stochastic parameters estimated with confidence using data generated from different experimental designs in human hepatocytes. Kinetic parameters were determined as species-dependent with the exception of Michaelis-Menten constant. In addition, a novel experimental and mechanistic modelling approach was developed using freshly isolated rat hepatocytes; method was subsequently applied to assess active uptake in plated cryopreserved human hepatocytes. Modelling and simulation was applied to optimise the experimental design conducted at a single concentration of 1 µM and over extended incubation time. The novel experimental protocol comprised of an uptake phase until initial equilibrium between cell and medium was reached, followed by a post-wash phase. The optimal post-wash medium, the stability of uptake transporters over time and the cell loss throughout the experiment were investigated. When required, the mechanistic compartmental model was extended to assess metabolism (repaglinide, cerivastatin, glibenclamide and rosiglitazone). The analysis of generated data allowed a good definition of the cellular concentration-time profiles of selected drugs and estimation of the kinetic parameters with precision. The decline in post-wash cellular concentrations was related to permeability properties, extent of intracellular binding and ratio of active clearance over passive diffusion of drugs investigated. Active process was a major contributor to the total uptake, ranging from 58-99.5 % for cerivastatin and fexofenadine in rat and from 62 % to 99.6 % for repaglinide and valsartan in human hepatocytes (FOP). In addition, the application of the novel design allowed clear differentiation of the cellular processes and reduced previously observed correlation between CLpassive and fu,cell. In rat, CLactive was on average 4.8-fold greater compared to human hepatocytes. CLactive was independent of SLCO1B1 c.521 genotype for rosuvastatin, valsartan and repaglinide (data from 4 donors). Direct comparison to data previously reported in literature in the same donor of human hepatocytes (HU4199) showed significant improvement in prediction of hepatic intrinsic clearance and reduction in under-prediction trend and need for empirical scaling factors for model drugs repaglinide, valsartan and rosuvastatin.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:701118 |
| Date | January 2016 |
| Creators | Badee, Justine Marine |
| Contributors | Galetin, Aleksandra ; Rostami-Hochaghan, Amin |
| Publisher | University of Manchester |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.1625 seconds