Skin sensitization, false positives and false negatives: experience with guinea pig assays

The advent of the local lymph node assay (LLNA), and efforts to develop in vitro alternatives for the identification of skin sensitizing chemicals has focused attention on the issue of false positive and false negative results. In essence, the question becomes 'what is the gold standard?' In this context, attention has focused primarily on the LLNA as this is now the preferred assay for skin sensitization testing. However, for many years prior to introduction of the LLNA, the guinea pig maximization test and the occluded patch test of Buehler were the methods of choice. In order to encourage a more informed dialogue about the relative performance, accuracy and applicability of the LLNA and guinea pig tests, we have here considered the extent to which guinea pig methods were themselves subject to false positives and negative results. We describe and discuss here well-characterized examples of instances where both false negatives (including abietic acid and eugenol) or false positives (including vanillin and sulfanilic acid) have been recorded in guinea pig tests. These and other examples are discussed with particular reference to the fabrication of a gold standard dataset that is required for the validation of in vitro alternatives.