TABLE 4

Examples of PBPK modeling impacting mechanistic understanding of clinical observations in GlaxoSmithKline (GSK)

DrugScenarioSimulationImpact
GSK28CYP3A4 substrateSimulated effect of ketoconazoleMechanistic understanding of elimination pathways.
GSK29Convulsions observed in dog (toxicology species)Predicted drug concentrations in human and dog brainTranslation of safety margins for convulsions, discharging risk in human.
GSK30CYP2D6 metabolism dependent inhibitorSimulated perpetrator drug interaction with dextromethorphan as victim drugMechanistic understanding of observed drug interaction with dextromethorphan.
GSK31PXR up-regulationConfirmed perpetrator drug interaction with midazolamMechanistic understanding of observed drug interaction and extrapolation to other cytokines.
GSK32UGT substrateSimulated impact of enterohepatic recycling on pharmacokinetic profiles and potential impact of reducing EHC with antibioticsMechanistic possibility for reduced exposure observed when dosing with comedications.
GSK33CYP3A4 substrate, inhaled doseSimulated ketoconazole drug interaction (as less drug interaction observed than expected based on metabolism by CYP3A4)Mechanistic understanding as redefining the inhaled dose enabled an accurate simulation of the observed drug interaction.
GSK34Biphasic half-life, long terminal half-lifeSimulated pharmacokinetic profilesMechanistic understanding of long half-life.
GSK35Metabolite is a CYP3A4 inhibitorSimulated contribution of a more potent CYP inhibitor metabolite to a victim drug interactionUnderstanding of CYP3A4 MDI mechanism.
GSK36CYP3A4 substrateSimulated drug interaction with ketoconazole (200 and 400 mg doses)Understanding of observed clinical ketoconazole data.
GSK37CYP3A4 substrateSimulated drug interactions with ketoconazole, erythromycin, and ritonavirUnderstanding of observed clinical data with ketoconazole and erythromycin; a further study with ritonavir was not recommended.
GSK38CYP3A4 inducer and CYP3A4 substrateSimulated induction (using midazolam as a clinical probe) and autoinduction using a perpetrator dose range of 60–160 mgMechanistic understanding of observed midazolam drug interaction data and reduced exposures of GSK38 after repeat dosing.
Case 5
GSK39CYP3A4 inducerSimulated induction using midazolam as a clinical probePredicted drug interaction was in line with the observed data.
GSK40CYP3A4/PGP substrateMade victim drug predictions with Kaletra, erythromycin, fluconazole, itraconazole, and verapamilMechanistic understanding of the clinical drug interactions observed with Kaletra and ketoconazole. Validated a microdose victim drug interaction approach for a backup drug.
GSK41CYP3A4 inhibitorMade perpetrator drug predictions with midazolam and simvastatinMechanistic understanding of clinically observed DDI with midazolam.
  • EHC, enterohepatic circulation; PXR, pregnane X receptor; MDI, metabolism dependent inhibition; UGT, UDP-glucuronosyltransferase.