Lia Liefaard1, Chao Chen. 1. Clinical Pharmacology Modelling and Simulation, *GlaxoSmithKline, Stevenage, Hertfordshire; and †GlaxoSmithKline, Uxbridge, Middlesex, United Kingdom.
Abstract
BACKGROUND: Dose individualization can reduce variability in exposure. The objective of this work was to quantify, through pharmacokinetic (PK) simulation, the potential for reducing the variability in exposure by dose individualization for a drug with moderate PK variability between subjects and between occasions within a subject, and a narrow therapeutic window. METHODS: Using a population PK model that includes between-subject and between-occasion variability for apparent clearance, individual PK profiles in a trial of 300 subjects after a test dose were simulated. From the simulated data, datasets were created mimicking various sampling regimens (from single predose sample to full profile samples over 12 hours) on 1 or more occasions (1, 2, 3, 5, or 10 visits). Using these datasets, individual apparent clearance values were estimated, which were then used to calculate an individualized dose for a predefined target area under the concentration-time curve (AUC), based on the available formulation strengths. The proportion of people whose mean AUC was within a predefined therapeutic AUC range was calculated for the test (before) and the individualized dose (after), and compared between the different sampling scenarios. RESULTS: The maximum increase in proportion of subjects with an AUC within the range was 20%. To achieve this benefit, PK samples over 4 hours from 100 dosing occasions were required. As a result of the dose adjustment, the AUC of 7.3% of the subjects moved from inside the therapeutic range to outside of the range. CONCLUSIONS: This work shows how modeling and simulation can help assess the benefit and risk of dose individualization for a compound with variability between subjects and between occasions. The framework can be applied to similar situations with a defined set of conditions (eg, therapeutic window, tablet strengths, and PK and/or pharmacodynamic sampling scheme) to inform dose change and to assess the utility of dose individualization against certain success criteria.
BACKGROUND: Dose individualization can reduce variability in exposure. The objective of this work was to quantify, through pharmacokinetic (PK) simulation, the potential for reducing the variability in exposure by dose individualization for a drug with moderate PK variability between subjects and between occasions within a subject, and a narrow therapeutic window. METHODS: Using a population PK model that includes between-subject and between-occasion variability for apparent clearance, individual PK profiles in a trial of 300 subjects after a test dose were simulated. From the simulated data, datasets were created mimicking various sampling regimens (from single predose sample to full profile samples over 12 hours) on 1 or more occasions (1, 2, 3, 5, or 10 visits). Using these datasets, individual apparent clearance values were estimated, which were then used to calculate an individualized dose for a predefined target area under the concentration-time curve (AUC), based on the available formulation strengths. The proportion of people whose mean AUC was within a predefined therapeutic AUC range was calculated for the test (before) and the individualized dose (after), and compared between the different sampling scenarios. RESULTS: The maximum increase in proportion of subjects with an AUC within the range was 20%. To achieve this benefit, PK samples over 4 hours from 100 dosing occasions were required. As a result of the dose adjustment, the AUC of 7.3% of the subjects moved from inside the therapeutic range to outside of the range. CONCLUSIONS: This work shows how modeling and simulation can help assess the benefit and risk of dose individualization for a compound with variability between subjects and between occasions. The framework can be applied to similar situations with a defined set of conditions (eg, therapeutic window, tablet strengths, and PK and/or pharmacodynamic sampling scheme) to inform dose change and to assess the utility of dose individualization against certain success criteria.