Use of QSAR and Physico Chemical Exclusion Rules to Predict Skin Irritation Potential

During the past 20 years, the EU legislation for the notification of chemicals has focused on new chemicals and, at the same time, has failed to cover the evaluation of existing chemicals in Europe. Therefore, in a new EU chemicals policy (REACH; Registration, Evaluation and Authorisation of Chemicals), the European Commission proposes to evaluate 30 000 existing chemicals within a period of 15 years. Data on the hazardous properties of chemicals must be provided by industry, which must also cover the costs if testing is required. There is agreement within the EU Commission that additional toxicity testing should, in the first place, rely on non-animal in-vitro tests, both for financial and animal welfare reasons. A realistic scenario, based on an in-depth discussion of potential toxicological developments and an optimized "tailor-made" testing strategy, shows that in order to meet the goals of the REACH policy, animal numbers may be significantly reduced, if industry were to use in-house data from toxicity testing (which is confidential), and if non-animal tests were used as well as information from quantitative structure-activity relationships (QSARs) applied to substance-tailored testing schemes.

An evaluation of a set of QSAR rules for predicting the absence of skin irritation and/or corrosion was recently conducted by Ingrid Gerner and colleagues from the BfR [22-24]. Scientists from the National Institute of Health in the Netherlands (RIVM) have critically evaluated the BfR approach on behalf of the European Chemicals Bureau (ECB) of the European Commission [25]. The evaluation of the rule-base on irritation and corrosion was developed by the BfR, the German Federal Institute for Risk Assessment [23, 24]. This rule-base predicts non-irritation and non-corrosion using physico-chemical cut-off values, defining general rules applicable to all substances and separate rules for special chemical classes of substances.

The evaluation on behalf of the ECB includes first, the compliance of the rule-base with the OECD principles on (Q)SARs; second the derivation of the (Q)SAR rules; and finally, the external validation of these rules, including an assessment of the suitability of the dataset used for validation.

The distribution of the training set data over the domains ofthe physico-chemical parameters used in the rule-base is visualized and analyzed. Recommendations are given for setting the cut-off values of the rules at a consequently "safe" level (not allowing for any exception to the rule in the training set), and for including a consistently calculated safety margin. Specific results of the analysis were that:

• the rule-base fulfils the OECD principles on (Q)SARs for the largest part;

• most rules cover all irritant/corrosive substances in the training set; however:

• some physico-chemical parameters have a limited predictive value:

- lipid solubility, as this is not a generally available parameter,

- vapor pressure, as the experimental data used for derivation of the rules is not conclusive, and

- melting point, as the cut-off values for this parameter were not set at a "safe" level, making predictions based on melting point less reliable.

An external validation of the set of rules using 201 new substances not present in the training set, showed 99.3% correct predictions of non-corrosivity, and 96.6% correct predictions of non-irritancy. These predictions would allow declassification as R34/R35 - corrosive for 28.4% of the chemicals, and R38 - irritant for 42.3% of the chemicals. These results would thus allow the waiving of skin irritation tests for at least 42.3% of the EU's new substance notifications. Four predictions were incorrect, however: for three of these, reasons could be given why the set of rules failed (two substances were misclassified based on melting point rules shown to be unreliable in our analysis), and/or how these incorrect predictions can be avoided in the future. The performance of the rules then increases to 100% correct predictions of non-corrosivity and 98.8% correct predictions of non-irritancy.

The results show that the evaluated rule-base is highly useful for regulatory purposes, as almost all OECD principles on QSARs are met, and the good predictivity could lead to the waiving of skin irritation tests for at least 42.3% of EU new substance notifications [25].

Swine Influenza

Swine Influenza

SWINE INFLUENZA frightening you? CONCERNED about the health implications? Coughs and Sneezes Spread Diseases! Stop The Swine Flu from Spreading. Follow the advice to keep your family and friends safe from this virus and not become another victim. These simple cost free guidelines will help you to protect yourself from the swine flu.

Get My Free Ebook


Post a comment