Thompson, C (2023) Development of Tools to Support the Application of Physiologically-Based Kinetic Modelling in Safety Assessment. Doctoral thesis, Liverpool John Moores University.

Preview

Text 2022ThompsonPhD.pdf - Published Version Available under License Creative Commons Attribution Non-commercial. Download (4MB) | Preview

Abstract

In safety assessment, one alternative to animal testing, that is increasingly being recognised as a suitable method for obtaining data, is read-across. This is the process by which information from one (source) chemical is used to infer information for another similar (target) chemical. However, determining similarity between chemicals is difficult as they cannot be considered as absolutely similar, only similar with respect to a given property. The research undertaken involved the development of two resources to address this problem. Within this thesis a resource is presented, wherein available physiologically-based kinetic (PBK) models have been collated and used in the development of a second resource, a KNIME workflow that can help to identify similar chemicals that have an existing PBK model. This enables PBK model information from source chemical(s) to be used in a read-across approach to help develop new PBK models for target chemicals. Initially, a systematic review of available PBK models in literature was undertaken; extraction of relevant information from the papers identified was captured in a Microsoft Excel spreadsheet which consisted of over 7,500 PBK models. The chemical space coverage of the chemicals with PBK models was compared to that of six datasets for different types of chemicals (drugs, cosmetics, botanicals, industrial chemicals, pesticides and food additives) in terms of their key physicochemical properties (molecular weight, hydrogen bond acceptors/donors, Slog P, TPSA, and Lipinski rule violations). The PBK model dataset that had been created in Excel was used as a resource in the development of a KNIME workflow that helps to identify similar chemicals to a target chemical. This KNIME workflow was applied to six chemicals, one from each of the different datasets, to assess its applicability across various industries. Two complete case studies were then undertaken (using the drug atenolol and the pesticide flumioxazin) in which new PBK models were built for these two target chemicals, using data obtained from source chemicals that had been identified by the workflow as being similar. PBK models for the source chemicals were initially reproduced, before being adapted and used as templates for the target chemicals. The performance of the new PBK models was assessed by comparing the output of these to existing data. The results demonstrated that a read-across approach could be successfully applied for developing new PBK models for data-poor chemicals. Information acquired from these PBK models can be used to support safety assessment of chemicals, reducing reliance on animal testing.

Item Type:	Thesis (Doctoral)
Uncontrolled Keywords:	PBK; read-across; physiologically-based kinetic model
Subjects:	R Medicine > RS Pharmacy and materia medica
Divisions:	Pharmacy & Biomolecular Sciences
SWORD Depositor:	A Symplectic
Date Deposited:	30 Jan 2023 09:13
Last Modified:	30 Jan 2023 09:14
DOI or ID number:	10.24377/LJMU.t.00018629
Supervisors:	Madden, J, Webb, S and Penson, P
URI:	https://researchonline.ljmu.ac.uk/id/eprint/18629

View Item