Skin Sensitization: Strategies for the Assessment and Management of Risk

D.A. Basketter


The British Journal of Dermatology. 2008;159(2):267-273. 

In This Article

Hazard Identification

Test guidelines for essentially all of the important toxicology assays are contained in the publications of the Organization for Economic Cooperation and Development (OECD). The test guideline for guinea pig predictive assays for skin sensitization is No. 406.[5] This contains instructions for the conduct and interpretation of the occluded patch test of Buehler[6] and for the guinea pig maximization test (GPMT) developed a few years later by Magnusson and Kligman.[7] Neither of these assays has been validated formally, but it is generally recognized that when performed well they successfully identify significant skin sensitizers.[8,9,10]

What does that expression mean - significant skin sensitizers? Globally, regulatory guidelines are set out in such a way that requires only chemicals which cause 15% or more guinea pigs to be sensitized in the Buehler test, or 30% or more to be sensitized in the GPMT, to be classified as skin sensitizers. In the EU, following the provision of the Dangerous Substances Directive, this then triggers labelling as 'R43: May cause sensitization by skin contact'.[11] Thus chemicals which are not labelled may either be true nonsensitizers or true skin sensitizers which are simply too weak to be classified.[12]

In more recent years, a murine test, the LLNA, was developed as an alternative to the guinea pig tests.[13,14,15,16] The method is shown in outline form in Figure 1. In the LLNA, sensitizing activity is measured as a function of proliferation in lymph nodes draining the site of chemical application, this being detected via tritiated thymidine incorporation. As with the guinea pig tests, a threshold was established, comparing proliferation in test vs. control lymph nodes.[17,18] This threshold, threefold or more proliferation in test vs. control, again permitted the discrimination of significant sensitizers from other chemicals. Following extensive trials and interlaboratory evaluations,[19] the LLNA was formally validated via the independent system set up for this purpose in the U.S.A.[20,21] and then subsequently formally endorsed in Europe.[22] The database for validation was formed from existing guinea pig and human data, the latter forming an important component of the validation considerations, at least in the U.S.A.[23]

Figure 1.

Overview of the local lymph node assay, in which a test substance is applied to the ear for 3days, followed on day 6 by the injection of tritiated thymidine into the tail vein, with subsequent removal and processing of the draining lymph nodes 3h later.

All predictive toxicology tests have their strengths and limitations, however. To a great extent it is the quality of our understanding of these which will determine how useful the test really is in practice. In the case of the guinea pig tests, for example, the quality of many subtle technical aspects of the conduct of the assay can impact substantially on the outcome, leading to false positives and false negatives.[10,12,24,25] These include how well the fur is removed, the vigour of the occlusion of the topical applications and the sensitivity of the subjective grading of the skin reactions at challenge. The impact of this can be seen in reports of the variability of the

The LLNA is also susceptible to technical variation, although here the simplicity of the protocol and the objective, quantitative nature of the readout mean that the method is much less susceptible to technical error; nevertheless, false positives and negatives do still occur.[29,30] Much has been written recently on this topic.[30,31,32] The key point to make here is that correct hazard identification is the fundamental on which all the subsequent risk assessment and risk management will be based, meaning that proper interpretation of hazard data, excluding both under- and over-interpretation, is vital. Under-interpretation (failure to recognize a skin sensitizer, or at least to underestimate its potency) may lead to avoidable skin disease; over-interpretation may lead to a devaluation of the whole risk assessment/management process and so in a real sense is arguably the greater potential error.

Of course, skin sensitizers can also be identified on the basis of clinical evidence.[11] Isolated episodes of ACD are not sufficient by themselves; typically what is necessary is evidence that the sensitizer can be identified as a problem in multiple dermatology clinics with a sufficient frequency to cause concern, always keeping in mind that frequency depends on exposure, which is not an intrinsic property of a substance. Thus a low frequency where there is heavy exposure might not lead to classification, whereas a higher frequency with lower exposure would be likely to trigger an R43 classification.[11]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: