Endpoint data, classifying

A9.6.4.2 Test data always take precedence over QSAR predications, providing the test data are valid, with QSARs used for filling data gaps for purposes of classification. Since the available QSARs are of varying reliability and application range, different restrictions apply for the prediction of each of these endpoints. Nevertheless, if a tested compound belongs to a chemical class or structure type (see above) for which there is some confidence in the predictive utility of the QSAR model, it is worthwhile to compare this prediction with the experimental data, as it is not unusual to use this approach to detect some of the experimental artefacts (volatilization, insufficient test duration to achieve equilibrium, and water solubility cut-off) in the measured data, which would mostly result in classifying substances as lower than actual toxicity. 

The ideal test to answer a safety assessment question should have an endpoint measurement that provides data such that dose-response relationships can be obtained. Furthermore, any criterion of effect must be sufficiently accurate in the sense that it can be used to reliably resolve the relative toxicity of two test chemicals that produce distinct yet similar responses (in terms of hazard to humans). In general, it may not be sufficient to classify test chemicals into generic toxicity categories. For instance, if a test chemical falls into an intermediate toxicity category but is borderline to the next, more severe toxicity category, it should be treated with greater concern than another test chemical that falls at the less toxic extreme of the... 

It is difficult to obtain a set of test substances that has consistent, high-quality toxicity data. The difficulty arises because many different schemes are used for measuring and classifying toxicity endpoints. The situation is made worse by the fact that there is no consistent source for the information that currently exists. This has meant that toxicity data used in validation studies have often come from many laboratories that have used different protocols and produced different kinds of data. 

The IRE assay has been assessed in several validation studies including the EC/HO study, the COLIPA study, the CTFA study and the IRAG evaluation (for review see 13). More recently, a retrospective evaluation study carried out by ICCVAM combined all existing information from previous studies and evaluated the usefulness of the assay to identify serious eye damage [54] as well as non-classified from classified chemicals [55]. However, further improvement and analyses were recommended before a statement on the scientific validity of the HET-CAM could be made. The reason for this is the fact that several endpoints and protocols for the IRE were applied and evaluated, each with insufficient data provided to make a sound conclusion [76]. 

The development of QSARs for any endpoint requires biological effect data to model. Without these data, no modehng is possible. The hmitations of the data to be modeled (both biological and physicochemical) must be appreciated by both the model developer and user. Criteria have been estabhshed to classify data for QSAR analysis according to their quahty (Cronin, 2005 Cronin and Schultz, 2003). For some biological effects, large coherent databases have been specifically created for the development of products and possibly QSAR modeling. Unfortunately, this has not been the case in the area of skin permeabihty, and the enthusiastic QSAR modeler is left with historical literatore data unless the modeler has a personal source of in-house data to model. 

The in vitro assay was based on HPTC seeded into uncoated multiwell plates that were cultivated for 3 days until the cells had formed a differentiated simple epithelium. As endpoint the expression levels of interleukin (IL)6 and IL8 were determined by qPCR. In the first analysis performed on the expression data (Li et al., 2013), a result was classified as positive if the drug-induced increase of at least one of these markers (IL6 or IL8) was above a certain threshold. When this thresholding procedure was applied, the median values (three batches of HPTC) of all major performance metrics... 

---