Physiologically Based Pharmacokinetic Modelling to Identify Physiological and Drug Parameters Driving Pharmacokinetics in Obese Individuals

Abstract

Background Obese individuals are often underrepresented in clinical trials, leading to a lack of dosing guidance. Objective This study aimed to investigate which physiological parameters and drug properties determine drug disposition changes in obese using our physiologically based pharmacokinetic (PBPK) framework, informed with obese population characteristics. Methods Simulations were performed for ten drugs with clinical data in obese (i.e., midazolam, triazolam, caffeine, chlorzoxazone, acetaminophen, lorazepam, propranolol, amikacin, tobramycin, and glimepiride). PBPK drug models were developed and verified first against clinical data in non-obese (body mass index (BMI) ≤ 30 kg/m 2 ) and subsequently in obese (BMI ≥ 30 kg/m 2 ) without changing any drug parameters. Additionally, the PBPK model was used to study the effect of obesity on the pharmacokinetic parameters by simulating drug disposition across BMI, starting from 20 up to 60 kg/m 2 . Results Predicted pharmacokinetic parameters were within 1.25-fold (71.5%), 1.5-fold (21.5%) and twofold (7%) of clinical data. On average, clearance increased by 1.6% per BMI unit up to 64% for a BMI of 60 kg/m 2 , which was explained by the increased hepatic and renal blood flows. Volume of distribution increased for all drugs up to threefold for a BMI of 60 kg/m 2 ; this change was driven by pK a for ionized drugs and logP for neutral and unionized drugs. C max decreased similarly across all drugs while t max remained unchanged. Conclusion Both physiological changes and drug properties impact drug pharmacokinetics in obese subjects. Clearance increases due to enhanced hepatic and renal blood flows. Volume of distribution is higher for all drugs, with differences among drugs depending on their pK a /logP. Extended author information available on the last page of the article Key PointsDrug pharmacokinetics in obese subjects (BMI ≥ 30 kg/m 2 ) were accurately predicted by our physiologically based pharmacokinetic framework informed with our previously developed obese population repository (i.e., 71.5% of predictions were within the 1.25-fold and none outside the twofold of clinical data).The model predicted an average 1.6% increase in clearance per BMI unit, which was explained by an increase in hepatic and renal blood flows. The volume of distribution increased on average up to threefold for a BMI of 60 kg/m 2 ; this change was driven by drug properties (pK a and logP). Dosing in obese subjects needs to take into account both physiological parameters and drug properties.

Is There a Need for a Dedicated Pharmacokinetic Trial for a Drug in Obese Populations? A Drug Prioritization Decision Tree Framework

“…Like CYP2E1, UGTs have shown a higher expression in people with obesity in some studies. Physiologically based pharmacokinetic modeling (PBPK) analysis of lorazepam indicated an 18% increase in UGT expression, with the increased enzyme expression in people with obesity confirmed using several other substrates 45 . Explanations for the increase in the activity of UGT include increased liver organ size and UGT expression in adipose tissue 38,46 .…”

“…Physiologically based pharmacokinetic modeling (PBPK) analysis of lorazepam indicated an 18% increase in UGT expression, with the increased enzyme expression in people with obesity confirmed using several other substrates. 45 Explanations for the increase in the activity of UGT include increased liver organ size and UGT expression in adipose tissue. 38,46 In a population pharmacokinetic analysis of garenoxacin, a primary UGT substrate, results showed that an obesity factor was a significant covariate on clearance (12.5% increase for the typical subject), leading to a significant increase in the clearance of patients who are obese when compared with patients of healthy weight.…”

Model‐Informed Approaches to Support Drug Development for Patients With Obesity: A Regulatory Perspective

“…47 Excessive fat deposition and chronic inflammation in the liver induced by obesity have reportedly caused a range of liver abnormalities. Specifically, hepatic blood flow was found to be increased in subjects with obesity as a result of increased cardiac output, 51,52 potentially leading to elevated hepatic elimination of drugs with a high extraction ratio (eg, propofol and fentanyl) following intravenous administration. 25,53,54 In addition, drugs characterized as obtaining a low to medium extraction ratio in subjects without obesity may shift to the category of a medium to high extraction ratio in the obese population due to the complex influence of obesity on hepatic blood flow, plasma protein binding, and intrinsic drug clearance.…”

“…For example, in the OSA drug development case mentioned above, PBPK models have been used for informing the effects of investigational drugs on gastric emptying time and the subsequent drug-drug interaction potential with orally administered concomitant medications in patients with OSA with obesity (BMI, 30 kg/m 2 or greater). 92,93 The second example is the integrated glucose-insulin (IGI) model, which is a mechanistic QSP model that quantifies the relationship between drug concentration, insulin lispro-aabc (Lyumiev), and postprandial glucose (PPG) response following administration of the insulin product. 94,95 The model describes the glucose absorption (intestinal transit through the stomach, duodenum, jejunum, and ileum), distribution, and elimination process in the body, which is regulated by a feedback relationship between insulin and glucose response.…”

“…In regulatory submissions, PBPK modeling has been proposed to assess the potential impact of obesity on the drug's PK parameters during early drug development. For example, in the OSA drug development case mentioned above, PBPK models have been used for informing the effects of investigational drugs on gastric emptying time and the subsequent drug–drug interaction potential with orally administered concomitant medications in patients with OSA with obesity (BMI, 30 kg/m 2 or greater) 92,93 . The second example is the integrated glucose‐insulin (IGI) model, which is a mechanistic QSP model that quantifies the relationship between drug concentration, insulin lispro‐aabc (Lyumiev), and postprandial glucose (PPG) response following administration of the insulin product 94,95 .…”