FS04.3 Hair dyes, prediction of sensitization potential with QSAR

Objectives:  To identify substances and tonnage data for use for hair dyes registered in Europe. To predict the sensitization potential of each substance and to rank the substances due to their sensitization potential. Further to group the substances in clusters based on their physical chemical properties with a cluster analysis. Methods:  The Inventory list of Cosmetics Ingredients (INCI), new regulation on cosmetics, tonnage data for use and Toxnet were used to identify and quantify the hair dyes. Salts were disregarded. A QSAR (Qualitative structure-activity relationship) model called TOPS-MODE, based on local lymph node assay (LLNA) data and physical chemical properties were used to predict the sensitization potential and make a cluster analysis. Results:  Out of 315 hair dye substances 229 meet the inclusion criteria. Most of the hair dye substances 75% were predicted to be strong to moderate sensitizers. Less were predicted to be weak 22% and only a small part 3% were predicted to be extremely weak or non sensitizing. The 8 most used hair dye substances were predicted to be strong to moderate sensitizers. Ppd is the most used hair dye allergy marker but some azodyes were predicted to be more potent than para-phenylenediamine (ppd). Conclusions:  Most hair dye substances are predicted to be strong to moderate sensitizers, which explain why some people gets hair dye dermatitis. A patch test series with potent, much used azodyes, might prove useful in diagnosing ppd-negative patients, with hair dye allergy. The cluster analysis grouped the substances which can be helpful choosing substances for clinical patch test.