In vitro metabolism studies to inform physiologically-based pharmacokinetic modelling of mefloquine, ritonavir and proguanil

Cloete, Cleavon K

In vitro metabolism studies to inform physiologically-based pharmacokinetic modelling of mefloquine, ritonavir and proguanil

Thesis / Dissertation

2024

Publisher

University of Cape Town

Department

Department of Chemistry

Faculty

Faculty of Science

Abstract

Clinically observed variability in drug exposure has various negative implications, ranging from an increased possibility of adverse drug reactions to treatment failure. A key contributor to this variability is the contribution of drug metabolism, substantially mediated by the cytochrome P450 (CYP) enzyme superfamily. Thus, a valuable step in assessing the contribution of drug metabolism to variability in drug exposure is the measurement of the kinetics of CYP-mediated metabolism. However, this in vitro data needs to be translated to whole body in vivo clearance, which is also impacted by the physicochemical properties of a drug and how the drug interacts with physiological environments in the body. One method of doing so is through the use of physiologically based pharmacokinetic (PBPK) modelling. By combining in vitro data with available knowledge on the physicochemical properties of a drug PBPK modelling provides a mechanistic understanding of drug exposure. As such, the in vitro metabolic characteristics of 3 commonly used drugs, namely mefloquine, proguanil and ritonavir, were measured. This was done by measuring the microsomal metabolism of these drugs in human liver microsomes, then determining the fraction of the drug metabolized by CYP (fm,CYP) isoforms and confirming this metabolic pattern using recombinant CYP enzymes. In the case of mefloquine, it was found that the drug was metabolized mainly by CYP3A and CYP1A2. While mefloquine metabolism by CYP3A is widely reported, this is the first report of its metabolism by CYP1A2. Proguanil was found to be metabolized by a number of enzymes, namely CYP1A2, CYP2D6, CYP2C19 and CYP3A. This confirms reported data for CYP2C19 and CYP3A, and provides new data for CYP1A2 and CYP2D6. Lastly, ritonavir was found to be metabolized by CYP2D6 and CYP3A. These determined fm,CYPs were then used, along with available data from literature, to develop PBPK models for mefloquine, proguanil and cycloguanil as a metabolite of proguanil using a middle-out approach with the use of Simcyp modelling software. These models were verified by comparing the pharmacokinetic (PK) profiles they simulated to various PK profiles found in literature. The verified models were then applied to various scenarios of interest. In the case of mefloquine, these applications showed that in vitro data does not accurately scale to in vivo clearance which excludes the possibility of doing a bottom-up model for this compound. The verified PBPK model built using this data demonstrates the expected drug- drug interaction between mefloquine and a CYP1A2 inhibitor, fluvoxamine. The model also shows the limited effect of population differences on mefloquine PK, including the reported finding that polymorphisms of CYP3A5 do not affect mefloquine PK. Disease state may therefore have a bigger effect on the variability in mefloquine PK than variability in metabolism does. The model application for proguanil focussed on the potential effects of various CYP genotypes on proguanil metabolism given the number of polymorphic CYPs found to be involved in its metabolism. The effect of CYP2C19 polymorphisms on proguanil and cycloguanil exposure could be clearly demonstrated. Variability in metabolism by other CYPs does not have as big an impact on proguanil exposure, likely because of their lower contributions to the metabolism of the drug. Finally, a previously developed and verified ritonavir model was used, in conjunction with the mefloquine model developed and verified here, to explore the use of ritonavir as an inducer and inhibitor of CYP3A activity. This showed the limited ability of ritonavir to induce mefloquine metabolism, with it being a more capable inhibitor. The ability of ritonavir to induce CYP1A2 activity was also explored using the available model by optimizing its CYP1A2 induction capabilities through parameter optimization. This showed the importance of taking this induction into account when conducting DDI studies between ritonavir and substrates of CYP1A2. Overall, this work addresses gaps in the understanding of the in vitro metabolism of mefloquine, proguanil and ritonavir, and demonstrates how this data can be combined with PBPK modelling to understand and predict PK variability.

Keywords

Chemistry

Reference:

Collections

PhD / Doctoral

Full item page