Mechanistic pharmacokinetic / pharmacodynamic modelling of acute inflammatory challenge models

  • Robert Michael Willis

Student thesis: Doctoral Thesis

Abstract

Systemic Lupus Erythematosus (SLE), is a chronic autoimmune disease that can affectmultiple organ systems. Survival rates have improved in recent years but SLE is stillcurrently “managed” rather than “cured” and traditional treatments can cause irreversibledamage, sometimes leading to death. Consequently, there is a need for new therapeuticagents that target specific components of the disease pathogenesis. This has provedchallenging since the exact aetiology of SLE is undefined, although most of the cells of theimmune system have been implicated in the disease process.The cytokine interferon (IFN) α has a key role in the early induction of the disease andongoing pathogenesis. Immune complexes of autoantigens and autoantibodies have beenfound in the sera of SLE patients and induce the production of IFNα by plasmacytoiddendritic cells (pDCs) through toll like receptors. IFNα binds to a specific receptor whichresults in the regulation of genes associated with the viral response (the IFN genesignature). This signature has been correlated with disease activity in SLE patientsreinforcing a role of IFNα in SLE. IFNα can also stimulate and inhibit a wide range ofcytokines. In SLE, a vicious circle is established with an ongoing production of IFNα frompDCs that maintains the autoimmune process.Lupus mouse models that mimic a human SLE like disease have previously been used toinvestigate new treatments for SLE, but have only delivered drugs that treat the symptomsrather than modifying the disease. An alternative approach is to use mechanism basedmodels to investigate the role of a particular pathway or disease mechanism. PK/PDmodelling of the data arising from challenge models can provide confidence that theefficacy observed in preclinical studies can be translated to the clinical setting and mayimprove the overall efficiency and success of drug discovery programmes. The IFNαpathway looks a promising target for new treatments for SLE and the induction of IFNα andrelated biomarkers appear to be translatable between preclinical and clinical species.Therefore this looks an appropriate pathway to investigate using mechanistic preclinicalPK/PD challenge models where the selected challenge agent would induces elements of theIFNα pathway under investigation.AimsThe aim of this project was to take a disease focused approach to select and investigate twopreclinical mechanistic acute in vivo PK/PD models. These models were selected based onthe hypothesis that the IFNα pathway has a key role in the pathogenesis of SLE.MethodsThis project investigated two preclinical mechanistic in vivo PK/PD challenge models withchallenge agents that stimulate different components of IFNα pathway. One was a highthroughput rodent model designed to enable screening of a number of compounds whilethe other was a primate model that was considered translatable to human.Mouse modelA mouse model was investigated that used the small molecule TLR7 agonist resiquimod toinduce IFNα production from pDCs. First, the pharmacokinetics of resiquimod weredetermined after iv infusion of 0.25 mg/kg. The relationship between IFNα response anddose was then determined over the range 0.04 to 4.8 mg/kg. Finally, a fixed dose of 0.4mg/kg was used to determine the reproducibility of the IFNα response between study days.Concentrations of resiquimod and IFNα were determined by LC-MSMS and ELISA analysisrespectively.The data were initially analysed by non-compartmental analysis to determine thepharmacokinetics of resiquimod. Statistical and power analysis were then applied todetermine the reproducibility of the model between study days and to understand thepotential utility of the model in the drug discovery. Finally, a population PK/PD analysis ofthe dose response relationship was conducted using a range of models that included anendogenous modulator function to describe the inhibition of excessive IFNα production.Primate modelA primate model was investigated that used recombinant human IFNα2b (INTRON A) toinvestigate the induction of a range of biomarkers downstream of the IFN receptor.Cynonologus monkeys received a subcutaneous administration of vehicle and IFNα2b atboth 3 MIU/kg and 10 MIU/kg. The pharmacokinetics of IFNα2b were determined and theinduction of 29 cytokines/chemokines, neopterin and body temperature was investigated.Concentrations of IFNα2b and other cytokines/chemokines were determined by Milliplex®magnetic bead panel, concentrations of neopterin were determined by ELISA analysis andbody temperature was determined using a rectal thermometer.The data were initially analysed by non-compartmental analysis to determine thepharmacokinetics of IFNα2b. A population PK/PD analysis of the IFNα2b induction of 6biomarkers was then conducted using an indirect response model with stimulation of input.ResultsMouse modelFollowing iv infusion, resiquimod had a blood clearance of 69 mL/min/kg, a volume ofdistribution at steady state of 2.4 L/kg and a terminal half-life of 0.5 h. There was low interanimal variability in pharmacokinetic parameters and similar results were obtained for ivinfusion and iv bolus administration. Linearity in pharmacokinetics was demonstrated overa 120 fold dose range.Following iv administration of resiquimod, IFNα concentrations were observed from 0.75 hpost dose, Cmax occurred at 1-2 h and the last measurable concentration was observed atapproximately 3 h. A bell shaped dose response curve was observed with the maximumresponse observed at 0.09 mg/kg. A dose of 0.4 mg/kg achieved reproducible responseresults across four study days and was recommended as the challenge dose for futurestudies. Power analysis demonstrated that the model could be used to investigate multiplecompounds at a single dose and the dose response of a single compound.Population PK/PD modelling was conducted in a sequential manner. The pharmacokineticsof resiquimod were described with a 1 compartment iv bolus model with IIV on clearance.The final PK/PD model comprised an indirect response model with stimulation of input, aneffect compartment and endogenous modulator function with fixed parameter estimatesfor Keo, the Hill co-efficient, and M50 and IIV on Keo, EC50 and Emax. The final model did notadequately predict the IFNα time profile in individual mice and highlights that increaseddata may be required to provide robust estimates of the induction and elimination phase ofthe IFNα response. The final model did characterise the dose response relationship butpredicted a reduction in response at higher doses of resiquimod when IIV was included onEmax rather than the modulator function.Primate modelFollowing sc administration, IFNα2b had an apparent blood clearance of 2.9 mL/min/kg, aapparent volume of distribution of 3.1 L/kg and a terminal half-life of 13 h. There was lowinter-animal variability in the systemic exposure (CV of 25-40%) and clearance (CV of 22-37%), however the volume and half-life demonstrated greater variability (CV of >100%)Linearity in pharmacokinetics was demonstrated over a 3 fold dose range. Thepharmacokinetics were broadly comparable to those previously reported in the literature.Following IFNα2b treatment induction of neopterin and the cytokines/chemokines IL15,IL1Ra, MCP1, IL6 and eotaxin was observed. There was a sub proportional increase in theconcentrations of biomarkers with the increase of IFNα2b dose from 3 MIU/kg to 10MIU/kg. An induction of body temperature was not observed.Population PK/PD modelling was conducted in a sequential manner. The pharmacokineticsof IFNα2b were described with a 2 compartment model with first order absorption and IIVon clearance and IOV on volume. A covariate analysis indicated there was no relationshipwith either clearance or volume with body weight. The final PK/PD model for all 6biomarkers comprised an indirect response model with stimulation of input and IIV onbaseline. In addition IIV on Emax was included for neopterin, IL15, IL6 and MCP1, on Kout forIL15 and EC50 for IL1Ra and eotaxin. High values for IIV were determined for IL6 and IL1Raand the parameters for MCP1 and IL6 demonstrated poor precision. The model did not givea robust prediction of EC50 for any biomarker or Emax for IL6, MCP1 and eotaxin due to theinvestigated doses giving a comparable response.ConclusionsMouse ModelThe pharmacokinetics of resiquimod were successfully determined in the mouse for thefirst time following both iv infusion and iv bolus administration of resiquimod. Theinvestigation of the dose response relationship delivered comparable data to thatpreviously reported in the literature whereby the greatest response was observed at lowdoses of resiquimod and further increase in dose results in an apparent reduction in theIFNα response to a plateau. A reproducible IFNα response can be achieved between micereceiving the same resiquimod treatment across multiple study days and suggests thatmodel that may be used in drug discovery to investigate therapeutics that have an actionon the IFNα pathway. Population PK/PD modelling with the incorporation of anendogenous modulator has highlighted the gaps in the data set and improved theunderstanding of the model that can be used to guide future PK/PD modelling efforts.Monkey ModelThe pharmacokinetics of IFNα2b demonstrated that at the doses investigated in this studythe receptor mediated clearance had become saturated and the clearance was driven byrenal elimination and catabolism. The induction of serum neopterin concentrations andcytokines/chemokines IL15, IL6, MCP1, eotaxin and IL1Ra, which have all been implicated inthe pathogenesis of SLE, was observed in the challenged primates. A less than proportionalincrease in biomarker concentrations was observed with the increase in dose indicatingthat the concentrations of IFNα2b associated with these doses may be near the top of theconcentration response curve. A population PK/PD model that describes the induction ofthe 6 biomarkers was developed. However, the model was limited by the small number ofdoses investigated. Due to the translatability of the biology between primate and humansthis project has delivered a PK/PD model with disease relevant endpoints that can be usedto screen compounds and potentially predict efficacious clinical doses.This project has successfully designed and validated two preclinical in vivo mechanisticPK/PD challenge models based on the hypothesis that the cytokine IFNα is central to thepathogenesis in SLE and that the mechanisms behind the induced response are translatablebetween species. This project has demonstrated that both models have potential in drugdiscovery to be used as tools to select the most appropriate compounds for progression tothe clinic and to predict efficacious doses.
Date of Award25 May 2017
Original languageEnglish
Awarding Institution
  • University Of Strathclyde
SponsorsUniversity of Strathclyde

Cite this

'