Florian Simon (1), Elodie Gautier-Veyret (2), Marylore Chenel (3), Léa Payen (4) Françoise Stanke-Labesque (2), Michel Tod (1)
(1) EA3738, Faculté de médecine de Lyon-Sud, Université de Lyon 1, Oullins, France, (2) Laboratoire de Pharmacologie, Pharmacogénétique et Toxicologie, Centre Hospitalier Universitaire des Alpes, Grenoble 38043, France, (3) Institut de recherches internationales Servier, Direction of clinical PK and pharmacometrics, Suresnes, France, (4) Laboratoire de biochimie-toxicologie, Centre hospitalier Lyon-Sud, Hospices civils de Lyon, Pierre Bénite, France
Objectives:
Systemic inflammation is known to impact drugs pharmacokinetics (PK) by different ways such as compromising metabolism and elimination of drugs. Plasma binding proteins levels are also impacted by inflammatory response resulting eventually in drug unbound fraction alteration [1]. Significant differences of PK profiles have been highlighted by clinical studies between patients with high or low systemic inflammation level [2]. These modifications could have clinical consequences and require dosing rate adaptations. In preliminary works, in vitro tests were conducted on intestinal microtissues in order to evaluate the impact of interleukin-6 exposure on different drug-metabolizing enzymes (DME) and drug-transporters expression and activity [3]. Results confirmed the trend already observed in several studies on hepatic cell lines: CYP2C19 and -3A4, majors DMEs, were strongly impacted. Then, a work was performed to relate, with physiologically-based PK (PBPK) modelling approach, inflammation with CYP3A4 metabolic clearance activity. A model was built and allow to predict midazolam, a CYP3A4 substrate, PK profile according to CRP level [4]. For this present study, voriconazole was chose as a substrate of CYP2C19 and -3A4.
The aim of this work was to build a PBPK model to describe the impact of inflammation on CYP2C19 activity and predict voriconazole PK profil in order of CRP level.
Methods:
Population model – Inflammatory effect calibration
PK data come from 91 hematological patients recipients of allogeneic hematopoietic stem cell transplantation [5, 6]. Patients receive intravenous or oral administration of voriconazole twice daily. Plasma samples were drawn 12 hours post-dose at steady-state, 315 trough concentrations were collected.
Model development was guided by the previously developed model from Dolton et al [7]. Two non-linear elimination pathways were used corresponding to CYP2C19 and CYP3A4. Fraction metabolized by each cytochrome, respectively 0.68 and 0.31 [8] and the ratio between Km2C19 and Km3A4 [9] were fixed. CYP2C19 genetic polymorphism analysis was performed and patients were assigned a CYP2C19 fractional activity (FA) with respect to the reference genotype [10]. Drug-drug interaction (DDI) with proton pump inhibitors (PPI) and glucocorticoids were taken into account using inhibition (IR) and induction potency factors (IC) on both cytochromes clearances [7]. The relation between CRP concentration and CYP3A4 was provided from our previous work on midazolam [4], Vmax3A4 was multiplied by kcrp = 0.2+0.8*exp(-ln2/25*[CRP]). Inter-individual variability (IIV) was described by an exponential model and the residual error by a combined proportional and additive error model. Data were analyzed using NONMEM 7.4. PsN was used for model run execution, Xpose and Pirana for post-processing analysis.
PBPK models – Inflammatory effect validation
In order to validate the k2C19 relation, 489 voriconazole trough concentrations from 34 patients were used [11]. Data were splited in four CRP groups (0-50, 50-100, 100-150 and >150 mg/L. A second validation was performed using omeprazole, another CYP2C19 substrate. Data of 12 rheumatoid arthritis patients taking 20 mg omeprazole (oral route), before and after administration of sirukumab (an anti-IL-6 monoclonal antibody) [12] were used. The inflammatory status and the CRP decreased after the antibody intake from 25.3 mg/L to 0.7 mg/L. The mean PK profiles and non-compartmental analysis parameters (Cmax, AUCinf and half-life) were provided.
PBPK models were built using PK-sim 8.0. DDIs, FAs for each CYP2C19 genotype and relationships between CRP concentrations and CYP P450s metabolic clearances according to the previous population model results, were processed with Mobi 8.0. Metabolic CYP P450s activities were described as saturable processes following Michaelis-Menten kinetics.
Models were validated using visual predictive checks (VPC). Simulations (N=1000) were performed by replacing [CRP] in k2C19 and k3A4 equations by the CRP corresponding value for each case. Median, 25th and 75th percentiles of trough concentrations in the four groups were compared to observations for voriconazole. Mean simulated PK profiles and parameters were compared the observed ones for omeprazole.
Results:
Different typical population and IIV values were taken from Dolton et al. and fixed in the population model: absorption constant ka = 0.53 h-1 IIV = 41.6%, biodisponibility factor F = 94.2% IIV = 36.7%, central volume of distribution Vc = 27.1L IIV = 83.4%, inter-compartmental clearance Q = 35.1L IIV = 37.4% and peripheral volume of distribution Vp = 127L IIV = 38.1%. Non-linear elimination parameters of CYP 2C19 were estimated: Vmax2C19 = 69.7 mg/h IIV = 9.70% and Km2C19 = 7.57 mg/L IIV = 31.9%. Vmax3A4 and Km3A4 were not estimated as Vmax3A4 = (0.31/0.68)*(Km3A4 / Km2C19)*Vmax2C19 and Km2C19=4*Km3A4
Estimation of PPIs inhibition potency factors improved the model. Typical values were IR2C19 = 0.31 and IR3A4 = 0.12 for esomeprazole and IR2C19 = 0.16 for pantoprazole. Glucocorticoids induction factors were fixed using covariates values from Dolton et al. [7] and FAs of CYP2C19 using values from Goutelle et al. [10]. CYP2C19 activity was related to CRP by multiplying Vmax2C19 by k2C19, , kcrp0 and alpha were estimated and the typical values were: kcrp0 = 0.49 and alpha = 0.086 L/mg. Adding all these covariates improve the objective function and the fold errors (= PRED/DV) of the model.
Using the population model result, CRP level was related to CYP P450s metabolic activities in Mobi by including k3A4 and k2C19 factors on metabolites formation rate equation. The VPCs were in good agreement with observations for the four voriconazole groups and for both omeprazole profiles.
Conclusions:
These findings suggest that voriconazole PK could be predicted depending on the level of CRP. This model may be useful for adapting the dosing rate of CYP3A4 and 2C19 substrates and to avoid toxicities in case of high inflammatory response. The approach will be extended in order to study substrates with high protein plasma binding in plasma. Effect of inflammation on DDI will also be studied.
References:
[1] Mayo PR, Skeith K, Russell AS, Jamali F. Decreased dromotropic response to verapamil despite pronounced increased drug concentration in rheumatoid arthritis. Br J Clin Pharmacol. 2000;50(6):605-13.
[2] Schmitt C, Kuhn B, Zhang X, Kivitz AJ, Grange S. Disease-drug-drug interaction involving tocilizumab and simvastatin in patients with rheumatoid arthritis. Clin Pharmacol Ther. 2011;89(5):735-40.
[3] Simon F, Garcia J, Guyot L, Guitton J, Vilchez G, Bardel C, et al. Impact of Interleukin-6 on Drug-Metabolizing Enzymes and Transporters in Intestinal Cells. AAPS J. 2019;22(1):16.
[4] Simon F et al PAGE 28 (2019) Abstr 9175 [www.page-meeting.org/?abstract=9175]
[5] Gautier-Veyret E, Bailly S, Fonrose X, Tonini J, Chevalier S, Thiebaut-Bertrand A, et al. Pharmacogenetics may influence the impact of inflammation on voriconazole trough concentrations. Pharmacogenomics. 2017;18(12):1119-23.
[6] Gautier-Veyret E, Truffot A, Bailly S, Fonrose X, Thiebaut-Bertrand A, Tonini J, et al. Inflammation is a potential risk factor of voriconazole overdose in hematological patients. Fundam Clin Pharmacol. 2019;33(2):232-8.
[7] Dolton MJ, Mikus G, Weiss J, Ray JE, McLachlan AJ. Understanding variability with voriconazole using a population pharmacokinetic approach: implications for optimal dosing. J Antimicrob Chemother. 2014;69(6):1633-41.
[8] https://www.ddi-predictor.org/
[9] Damle B, Varma MV, Wood N. Pharmacokinetics of voriconazole administered concomitantly with fluconazole and population-based simulation for sequential use. Antimicrob Agents Chemother. 2011;55(11):5172-7.
[10] Goutelle S, Bourguignon L, Bleyzac N, Berry J, Clavel-Grabit F, Tod M. In vivo quantitative prediction of the effect of gene polymorphisms and drug interactions on drug exposure for CYP2C19 substrates. AAPS J. 2013;15(2):415-26.
[11] Veringa A, Ter Avest M, Span LF, van den Heuvel ER, Touw DJ, Zijlstra JG, et al. Voriconazole metabolism is influenced by severe inflammation: a prospective study. J Antimicrob Chemother. 2017;72(1):261-7.
[12] Zhuang Y, de Vries DE, Xu Z, Marciniak SJ, Jr., Chen D, Leon F, et al. Evaluation of disease-mediated therapeutic protein-drug interactions between an anti-interleukin-6 monoclonal antibody (sirukumab) and cytochrome P450 activities in a phase 1 study in patients with rheumatoid arthritis using a cocktail approach. J Clin Pharmacol. 2015;55(12):1386-94.
Reference: PAGE () Abstr 9527 [www.page-meeting.org/?abstract=9527]
Poster: Oral: Drug/Disease Modelling