Investigating the state of physiologically based kinetic modelling practices and challenges associated with gaining regulatory acceptance of model applications

Abstract

Physiologically based kinetic (PBK) models are used widely throughout a number of working sectors, including academia and industry, to provide insight into the dosimetry related to observed adverse health effects in humans and other species. Use of these models has increased over the last several decades, especially in conjunction with emerging alternative methods to animal testing, such as in vitro studies and data-driven in silico quantitative-structure-activity-relationship (QSAR) predictions. Experimental information derived from these new approach methods can be used as input for model parameters and allows for increased confidence in models for chemicals that did not have in vivo data for model calibration. Despite significant advancements in good modelling practice (GMP) for model development and evaluation, there remains some reluctance among regulatory agencies to use such models during the risk assessment process. Here, the results of a survey disseminated to the modelling community are presented in order to inform the frequency of use and applications of PBK models in science and regulatory submission. Additionally, the survey was designed to identify a network of investigators involved in PBK modelling and knowledgeable of GMP so that they might be contacted in the future for peer review of PBK models, especially in regards to vetting the models to such a degree as to gain a greater acceptance for regulatory purposes.

Keywords: Alternatives methods; PBK models; Physiologically based kinetic models; Regulatory; Risk assessment; Toxicology.

Graphical abstract

Fig. 1

Number of papers published per year in the last 30 years. The search was conducted using PubMed on the 08th of April 2017, with key words including “PBPK model” only, or a set of keywords including the string “PBPK models OR PBBK models OR PBTK models OR PBK Models”. The year 2017 represents only papers published in the first 3 months.

Fig. 2

Geographic distribution of the questionnaire results, 93 respondents from 19 countries. Each country is labeled with the ISO 3166 International Standard Country Code, two letters, followed by the number of replies per country in brackets. The results were published and are publicly accessible at the following link http://bolegweb.geof.unizg.hr/questionaire/pbk/, to see individual results double click on a selected country. The color scale represents the lowest number of replies (black) to the highest number (light grey).

Fig. 3

A. Woking Sector of the 93 respondents to the survey and B. working expertise of the 93 respondents to the survey.

Fig. 4

The various language codes, software, or platforms used by survey participants to build PBK models for scientific and regulatory applications.

Fig. 5

Source of Chemical-Specific and Physiological parameters.

Fig. 6

Participants' suggestions of ADME properties that should be derived experimentally when developing PBK models, due to lack of data or information.

Fig. 7

A. Selected fields in which participants apply PBK models and B. selected applications of PBK models used by participants.