Mattia Berton (1,2), Sara Bettonte (1,2), Felix Stader (3), Manuel Battegay (1,2), Catia Marzolini (1,2,4)
(1) University Hospital Basel, Switzerland, (2) University of Basel, Switzerland, (3) Certara UK Limited, United Kingdom, (4) University of Liverpool, United Kingdom
Introduction: Obesity is a chronic and relapsing disease commonly defined by a body mass index (BMI) ≥ 30 kg/m2. The percentage of worldwide population affected by this condition is raising and represents today an important healthcare problem [1]. Obesity is characterised by anatomical, physiological, and biological remodelling which can lead to changes in drug pharmacokinetics [2]. This special population is often underrepresented in clinical trials leading to a lack of knowledge about drug kinetics and drug–drug interactions. Physiologically based pharmacokinetic (PBPK) modelling can overcome this clinical limitation, however detailed descriptions of the population characteristics are key to inform models.
Objectives: The purpose of this study was to develop and verify a population database for white obese individuals considering anatomical, physiological, and biological system parameters, including population variability, to inform a PBPK framework.
Methods: A systematic literature search was performed to collect anatomical, physiological, and biological parameters of interest such as organ weights, regional blood flows, plasma protein binding concentration and glomerular filtration rate. Articles were included if study subjects were 20 to 50 years old, were predominantly white, had a BMI greater than 18.5 kg/m2, and concurrent comorbidity was mild or deemed unlikely to affect the parameter of interest. Collected data were divided into a development and verification dataset based on the information reported in the study. Multiple regression analyses were performed to derive continuous mathematical functions describing the effect of BMI on system parameters. Covariates were considered significant when their p-value was lower than 0.01. Visual and numeric diagnostic (R2 and the Akaike’s information criterion) together with physiological plausibility were used to select the best mathematical function. Finally, the in-house PBPK model build in Matlab® [3] was implemented with the developed mathematical functions and to verify their predictivity three validation steps were performed: (1) system parameters of 1000 men and women with BMI ranging from 18.5 to 60 kg/m2 were predicted and compared against the independent verification dataset (2) the sum of organ weights and blood flows were checked to not exceed body weight and cardiac output (3) prediction of triazolam (CYP3A substrate, 0.5 mg single IV infusion) [4] and caffeine (CYP1A2 substrate, 162 mg single oral dose) [5-7] disposition in obese and non-obese were compared against clinically observed data.
Results: In total, 346 articles were screened of which 209 were included in the analysis. Overall information on organ weights were generally available in the literature up to BMI of 60 kg/m2, while measurements of regional blood flow and biological parameters were scarce and mostly available up to BMI of 40 kg/m2. Two important parameters for which there were no data in the literature were the microsomal proteins per gram liver (MPPGL) and the hepatocytes per gram liver (HPGL), which are essential hepatic scaling factors needed to extrapolate human clearance; to overcome the lack of data we assumed their value to be constant and we scaled them using only lean liver volume. Analysis of the data showed that a 3-fold increase in BMI, from 20 to 60 kg/m2, leads to an increase in the cardiac output (50%), in the liver weight (100%), in the kidney weight (60%), in both the kidney and liver absolute blood flows (50%), and in total adipose blood flow (160%). The simulation of the two drugs matched well the observed clinical data for both non-obese and obese individuals and the pharmacokinetic parameters were within 1.25-fold of clinically observed data. The observed versus predicted AUC0-∞ and half-life ratios between obese and non-obese adults were 1.20 vs 1.32 and 1.02 vs 0.86 for triazolam and 1.26 versus 1.35 and 0.75 versus 0.79 for caffeine, respectively.
Conclusion: The developed repository provides continuous functions describing anatomical, physiological, and biological changes induced by obesity. The equations together with the corresponding population variability can be implemented into existing PBPK frameworks and used to simulate drug disposition in obese. The predictive validity of the equations has been demonstrated by the good agreement between triazolam and caffeine predicted and observed clinical data.
References:
[1] World Health Organization. [cited 2021 November 4]; Available from: https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight.
[2] Brill, M.J., et al., Clin Pharmacokinet, 2012. 51(5): p. 277-304.
[3] Stader, F., et al., CPT Pharmacometrics Syst Pharmacol, 2019.
[4] Derry, C.L., et al., J Clin Psychopharmacol, 1995. 15(3): p. 197-205.
[5] Abernethy, D.R., E.L. Todd, and J.B. Schwartz, Br J Clin Pharmacol, 1985. 20(1): p. 61-6.
[6] Kaplan, G.B., et al., J Clin Pharmacol, 1997. 37(8): p. 693-703.
[7] Cysneiros, R.M., et al., Clin Pharmacol Ther, 2007. 82(1): p. 54-62.
Reference: PAGE 30 (2022) Abstr 10013 [www.page-meeting.org/?abstract=10013]
Poster: Drug/Disease Modelling - Other Topics