Regulatory agencies reluctant to use mathematical models of organisms

Regulatory agencies remain reluctant to use physiologically based kinetic (PBK) models, according to an expert survey by the European Commission’s Joint Research Centre (JRC). Widely used in industry and academia, PBK mathematical models describe how chemicals pass through the body, which is represented as a series of interconnected compartments. The models are becoming increasingly popular, particularly for use with non-animal test results and quantitative structure activity relationship (Qsar) predictions. For example, they are often used to help extrapolate from in vitro results to the in vivo situation. Of 93 people that filled out an online survey, 27 described using PBK models for submitting dossiers, opinions or risk assessments to regulatory agencies. Together with scientists at the US Environmental Protection Agency’s (EPA’s) Oak Ridge Institute and Slovakian consultancy Klimeto, the JRC team has summarised the PBK situation and come up with recommendations to boost regulatory acceptance. A lack of standardised good modelling practice (GMP) guidelines, together with a wide range of computing platforms, may be the “greatest obstacles” for PBK model use in regulatory risk assessment, they write in Regulatory Toxicology and Pharmacology. Establishing GMP throughout the modelling community could lead to a “standardised checklist” to evaluate the quality of a PBK model, which would help regulators, the authors say.

Challenges of newer PBK models

PBK models are often developed to answer specific scientific questions related to certain chemicals. However, when the purpose is to use a PBK model to support regulatory decision-making, the models require careful vetting of parameter values, model code and structure, they add. Developing new models can be labour intensive and data hungry. The researchers warn that newer PBK models bring “additional challenges” for risk assessors attempting to review them for regulatory decision making. When there are no experimental in vivo data to compare model predictions with, regulatory agencies are likely to remain reluctant to use models in their chemical safety assessments, they say. Finally, the report suggests that the next generation of PBK models could be developed solely using data from in vitro and in silico methods.

The JRC and EPA are co-leading an OECD project to harmonise characterisation, validation and reporting of PBK models. Further information is available at:

Chemical Watch, 30 November 2017 ; http://chemicalwatch.com