You are here:
Evaluating a Skin Sensitization Model and Examining Common Assumptions of Skin Sensitizers (QSAR conference)
Citation:
Fitzpatrick, J., D. Roberts, AND G. Tier. Evaluating a Skin Sensitization Model and Examining Common Assumptions of Skin Sensitizers (QSAR conference). Presented at 17th International Conference on QSAR in Environmental and Health Sciences, Miami, FL, June 13 - 17, 2016.
Impact/Purpose:
Presentation at QSAR conference on Evaluation of the existing skin sensitization model and uncovering new sensitization data.
Description:
Skin sensitization is an adverse outcome that has been well studied over many decades. Knowledge of the mechanism of action was recently summarized using the Adverse Outcome Pathway (AOP) framework as part of the OECD work programme (OECD, 2012). Currently there is a strong focus on how AOPs can be applied for different regulatory purposes including the development and application of Integrated Approaches to Testing and Assessment (IATA). One example is an Integrated Testing Strategy developed by Jaworska et al (2013) known as ITS-2 which was derived using a Bayesian network and which relied upon outcomes from different in vitro, in chemico and in silico approaches that mapped onto the events within the AOP. Here we assessed the performance of ITS-2 model using cross validation approaches. Re-training the network using Protein binding alerts as extracted from the OECD Toolbox in place of predictions from TIssue MEtabolism Simulation for Skin sensitization (TIMES-SS), a commercial expert system resulted in a comparable predictive performance. Skin penetration is denoted as the first event within the AOP since there is a long standing belief that substances that cannot penetrate through the skin will not be sensitizers. Physicochemical parameters such as LogKow and MW which are typical inputs to estimate skin penetration (specifically the skin permeability constant, Kp) have been used to propose thresholds for skin sensitization. Namely a sensitizer would have a MW>500 and a LogKow>1. A large dataset of 1155 substances that had been evaluated for their skin sensitization potential together with available measured LogKow values was compiled from the REACH dissemination database to investigate whether these thresholds were a myth or a reality. There were 176 compounds with a MW > 500 and 31 of these were sensitizers. This finding complemented that of Roberts et al. (2012) who examined the training set within TIMES-SS and found 13 compounds above a MW of 500 of which 5 were sensitizers. Comparing the incidence of sensitizers above and below a threshold of LogKow >1 showed no significant difference, 24% of substances below a LogKow of 1 were sensitizing compared with 31% above a LogKow of 1. These results support the notion that thresholds for LogKow and MW as surrogates for skin penetration are not helpful in predicting skin sensitization and cast doubt on the belief that penetration is even a relevant consideration for skin sensitization potential.Disclaimer: The views expressed are those of the authors and do not necessarily reflect the views or policies of the US Environmental Protection Agency.