Felix Stader (1,2,3), Hannah Kinvig (4), Melissa A. Penny (2,3), Manuel Battegay (1,3), Marco Siccardi (4) & Catia Marzolini (1,3)
(1) University Hospital Basel, Basel, Switzerland / (2) Swiss Tropical and Public Health Institute, Basel, Switzerland / (3) University of Basel, Basel, Switzerland / (4) Institute of Translational Medicine, University of Liverpool, Liverpool, UK
Objectives:
Aging is characterized by physiological and biological changes, which can affect drug pharmacokinetics. Even though the use of medications is highly prevalent, clinical studies are rarely conducted in the elderly. To overcome the sparse clinical data, physiologically based pharmacokinetic (PBPK) modelling was utilized to perform virtual clinical trials across the entire adult lifespan to investigate the impact of adult age on drug pharmacokinetics.
Methods:
A whole-body PBPK model constructed in Matlab® 2017a was used [1]. Healthy, virtual individuals aged 20 to 99 years were generated considering age-dependent changes of anatomical and physiological parameters [2]. A structured literature search was performed to identify drugs whose PK have been studied in the elderly to validate the simulations with clinical data. Selected drugs included midazolam, metoprolol, amlodipine, rivaroxaban, repaglinide, atorvastatin, rosuvastatin and lisinopril. Input drug parameters were obtained from existing validated PBPK models, except for lisinopril [1, 3-8]. Tissue distribution of the amlodipine model was modified to be used in a whole-body PBPK model based on the observed volume of distribution from an iv study in healthy men [9]. Active hepatic drug transport was included in the repaglinide PBPK model based on published in vitro data [10]. The lisinopril PBPK model was developed combining published in vitro data (bottom-up approach) with available clinical clearance data (top-down approach).
PBPK models were verified in young adults (20-50 years) following the best practice approach [11]. After successful prediction in young adults, judged by visual inspection of concentration-time profiles and prediction of PK parameters within 2-fold of the observed data, we carried out simulations in elderly adults (>65 years) without any modification to the drug parameters. Simulations were matched as closely as possible to the published observed studies in terms of demographics, dosing regimen and number of subjects with 10 trials x n virtual subjects being simulated for each drug.
The final PBPK models were utilised to predict drug pharmacokinetics from 20 to 99 years in 500 virtual subjects (proportion of women: 0.5) in five years steps. The analysed PK parameters (Maximal concentration: Cmax; time to maximal concentration: tmax; area under the curve: AUC; clearance: CL; volume of distribution: Vd; elimination half-life: t1/2) were normalised to the youngest investigated age group (20-24 years).
Results:
The simulation of all eight drugs matched well the observed clinical data for young (20-50 years) and elderly (>65 years) adults. The predicted AUC-ratio elderly:young adults was in close agreement to observed clinical data for midazolam (1.64 vs 1.96), metoprolol (1.20 vs 0.97), lisinopril (1.18 vs 1.24), amlodipine (1.31 vs 1.38), rivaroxaban (1.52 vs 1.52), repaglinide (1.62 vs 1.79), atorvastatin (1.32 vs 1.38) and rosuvastatin (1.24 vs 1.03). All other PK parameters were predicted within 2-fold of the observed data in young and elderly adults.
After successful model verification, the impact of adult age on the pharmacokinetics of the eight investigated drugs was examined. Cmax, tmax and Vd were independent of adult age. In contrast, AUC and t1/2 showed on average a progressive linear increase of 1.0% and 0.8% per year with age-related changes being more than expected from interindividual variability defined as the 1.25-fold interval from the age of 55 years. Accordingly, CL decreased linearly with a maximum 3.0-fold difference compared to young adults. Importantly, age-related changes for all investigated parameters were independent of gender.
Conclusions:
On a general rule, this study demonstrates that drug elimination rather than absorption or distribution drives age-related drug exposure changes in the elderly. For the first time, this study provides the scientific fundament for the 25-50% dose reduction used empirically by clinicians to treat elderly individuals [12] and shows principally that this dose reduction depends on the age-related decrease of liver weight, hepatic and renal blood flow as well as glomerular filtration rate and is independent of the drug. However, it needs to be emphasized that pharmacodynamic alterations and the presence of comorbidities should be considered when prescribing in the elderly.
References:
[1] Stader F, Penny MA, Siccardi M, & Marzolini C. A comprehensive framework for physiologically based pharmacokinetic modelling in Matlab®. CPT: Pharmacometrics & Systems Pharmacology, 2019. [Epub ahead of print].
[2] Stader F, Siccardi M, Battegay M, Kinvig H, Penny MA, & Marzolini C. Repository describing an aging population to inform physiologically based pharmacokinetic models considering anatomical, physiological, and biological age-dependent changes. Clinical Pharmacokinetics, 2018. [Epub ahead of print].
[3] Rowland-Yeo K, Walsky R, Jamei M, Rostami-Hodjegan A, & Tucker G. Prediction of time-dependent CYP3A4 drug–drug interactions by physiologically based pharmacokinetic modelling: impact of inactivation parameters and enzyme turnover. European Journal of Pharmaceutical Sciences, 2011. 43(3): 160-173.
[4] Chetty M, Rose RH, Abduljalil K, Patel N, Lu G, Cain T, Jamei M, & Rostami-Hodjegan A. Applications of linking PBPK and PD models to predict the impact of genotypic variability, formulation differences, differences in target binding capacity and target site drug concentrations on drug responses and variability. Frontiers in Pharmacology, 2014. 5(258): 1-14.
[5] Mukherjee D, Zha J, Menon RM, & Shebley M. Guiding dose adjustment of amlodipine after co-administration with ritonavir containing regimens using a physiologically-based pharmacokinetic/pharmacodynamic model. Journal of Pharmacokinetics and Pharmacodynamics, 2018. 45(3): 443-456.
[6] Marzolini C, Rajoli R, Battegay M, Elzi L, Back D, & Siccardi M. Physiologically based pharmacokinetic modeling to predict drug–drug interactions with efavirenz involving simultaneous inducing and inhibitory effects on cytochromes. Clinical Pharmacokinetics, 2017. 56(4): 409-420.
[7] Zhang T. Physiologically based pharmacokinetic modeling of disposition and drug–drug interactions for atorvastatin and its metabolites. European Journal of Pharmaceutical Sciences, 2015. 77: 216-229.
[8] Jamei M, Bajot F, Neuhoff S, Barter Z, Yang J, Rostami-Hodjegan A, & Rowland-Yeo K. A mechanistic framework for in vitro–in vivo extrapolation of liver membrane transporters: prediction of drug–drug interaction between rosuvastatin and cyclosporine. Clinical Pharmacokinetics, 2014. 53(1): 73-87.
[9] Faulkner J, McGibney D, Chasseaud L, Perry J, & Taylor I. The pharmacokinetics of amlodipine in healthy volunteers after single intravenous and oral doses and after 14 repeated oral doses given once daily. British Journal of Clinical Pharmacology, 1986. 22(1): 21-25.
[10] Varma MV, Lai Y, Kimoto E, Goosen TC, El-Kattan AF, & Kumar V. Mechanistic modeling to predict the transporter-and enzyme-mediated drug-drug interactions of repaglinide. Pharmaceutical Research, 2013. 30(4): 1188-1199.
[11] Ke A, Barter Z, Rowland-Yeo K, & Almond L. Towards a best practice approach in PBPK modeling: case example of developing a unified efavirenz model accounting for induction of CYPs 3A4 and 2B6. CPT: Pharmacometrics & Systems Pharmacology, 2016. 5(7): 367-376.
[12] Vass M & Hendriksen C. Medication for older people. Zeitschrift für Gerontologie und Geriatrie, 2005. 38(3): 190-195.
Reference: PAGE 28 (2019) Abstr 8891 [www.page-meeting.org/?abstract=8891]
Poster: Drug/Disease Modelling - Other Topics