Quantitative risk assessments for dermal sensitization to fragrance ingredients: The utility of LLNA data in the weight of evidence approach to identifying thresholds

Abstract

Historical human data from human repeated insult patch tests (HRIPT) and/or maximization tests (HMT) are available for fragrance ingredients used in consumer products. When conducting a Quantitative Risk Assessment for the induction of dermal sensitization, these data add to the overall weight of evidence approach used to determine potency. Contemporary human studies are not conducted to determine hazard; rather they confirm the lack of dermal sensitization at an exposure level identified as a NOEL in an animal model. One model commonly employed is the murine Local Lymph Node Assay (LLNA). In addition to identifying dermal sensitization hazards, the LLNA provides a quantitative measure of relative skin sensitizing potency. These potency estimates are based on interpolation of the dose response data, yielding an estimated concentration (EC3) required to elicit a positive response. In the present study, the EC3 values of 57 fragrance materials were compared to those derived from human NOELs for induction determined by historic confirmatory HRIPT and/or HMAX. The human NOELs and EC3 values were converted to their dose per unit area (μg/cm2) equivalents to allow for direct comparison. A good correlation existed between the EC3 values and the human NOELs for induction. The EC3 values were observed to predict and in some cases under predict the human NOELs for approximately ~80% of the materials tested. The results from this analysis demonstrate the utility of incorporating the EC3 value into the QRA approach. However, the lack of correlation for several materials highlights the importance of conducting a confirmatory HRIPT. Introduction When conducting a Quantitative Risk Assessment (QRA) for the induction of dermal sensitization, human data from repeated insult patch tests (HRIPT) and/or historical maximization tests (HMT) add to the overall weight of evidence approach used to determine potency (Api et al., 2008). Current practice is to conduct an HRIPT to confirm a predicted dermal sensitization no-effect level (NOEL) from animal testing. The Murine Local Lymph Node Assay (LLNA) is currently the most commonly utilized animal model to determine hazard and potency. The utility of combining LLNA data with knowledge gained from human studies in a weight of evidence approach applied to the QRA for dermal sensitization is considered below. Human Sensitization Test A human sensitization test is not used to determine hazard. The test is not used as a predictive method nor is it used on substances with unknown dermal sensitization potential. It is a test to confirm the lack of dermal sensitization at an exposure level which was identified as a NOEL in an animal model or derived as a likely NOEL from quantitative structure-activity relationships. Human patch testing methodology has evolved over more than 50 years. The test most typically conducted is the human repeated insult patch test (HRIPT) (McNamee et al., 2008). The HRIPT is generally performed in 100 subjects by conducting a total of nine 24-hour occluded applications over 3-weeks with test material and appropriate controls followed by a 2-week rest period. A single 24-hour challenge application is then made to a naive site with the same materials. Observations at challenge coupled with the patterns of reactivity observed during induction provide the basis for an interpretation of potential contact allergens. The HMT is no longer used by RIFM. Historically, it was conducted on 25 human subjects by utilizing 5 alternate day 48-hour occluded induction applications of test material and appropriate controls. Following a ten to fourteen-day rest period 48-hour challenge applications are made to naive sites. Patches may be made with and without pre-treatment of sodium lauryl sulfate depending upon the inherent irritancy of the test material. The HRIPT was designed and refined to exaggerate normal, realistic use conditions. This focus makes the HRIPT more relevant, when compared to the HMT, to the identification of a threshold for induction under foreseeable usage scenarios. Local Lymph Node Assay Data In recent years, the LLNA has been increasingly used for hazard assessment (Kimber et al., 1992; 1994). The LLNA measures events occurring after initial exposure to chemical sensitizers, specifically activity is measured as a function of lymphocyte proliferative responses induced in lymph nodes draining the site of topical exposure to the test chemical. In addition to its application as a method to identify potential contact allergens, the LLNA presents the opportunity for the objective and quantitative measure of relative skin sensitizing potency (Basketter et al., 2000). This potency estimate takes the form of an EC3 value—the concentration required to stimulate a three fold increase in lymphocyte proliferative responses. LLNA potency data were gathered from both published and unpublished reports for each material. Analysis The available data for 57 fragrance materials is presented in Table 1. The data show that for 40/57 of the fragrance ingredients reviewed, there is a very good correlation between the EC3 value from the LLNA and the NOEL in confirmatory human dermal sensitization tests. It should be pointed out that for 12 of the 40 materials, a true no effect level has not been determined in humans. The no effect level reported is the maximum tested concentration. For the remaining 17 materials, the correlation is less predictive. Of these 17 materials, the data for 5 (α-amylcinnamaldehyde, benzyl salicylate, α-hexylcinnamaldehyde, heyxl salicylate, and α–isomethylionone) reveal that the LLNA EC3 value is an order of magnitude greater than the maximum tested NOEL in humans (the NOEL reported reflects the highest concentration tested, not the highest achievable). The absence of significant clinically relevant positive reactions in dermatology clinics provides support for these data (see Table 2). For 12 materials (α-amylcinnamyl alcohol, benzaldehyde, benzyl alcohol, cinnamyl nitrile, coumarin, t-2-hexenal, isocyclogeraniol menthadiene-7-methyl formate, 6-methyl-3, 5-heptadien-2-one, methyl 2-nonynoate, 1-octen-3-yl acetate and treemoss) the data show that the EC3 value overestimates the NOEL in confirmatory human tests. Of these 12 materials only 4 (benzaldehyde, t-2-hexenal, and methyl 2-nonynoate) showed that the human no effect level was an order of magnitude lower than the LLNA EC3 value. The data for the remaining 8 materials showed that the human no effect level was lower than the LLNA EC value but the data are equivocal. None of the LLNAs were conducted up to 100%. Further there is no lowest observed effect level for α-amylcinnamyl alcohol data so it may be possible that the LLNA EC3 value and the human no effect level could be similar. These data illustrate the importance of conducting a confirmatory human sensitization test. There are seven essential oils that were reviewed and with the exception of treemoss the correlation between the LLNA EC3 value and the human data re good. These data are similar to the results found with other essential oils (Lalko and Api, 2006). References Api, A.M. et al., 2008. Regulatory Toxicology and Pharmacology, 52(1):3-23. Basketter, D.A., et al., 2000. Contact Dermatitis 42, 344-348. Kimber, I., Basketter, D., 1992. Food and Chemical Toxicology 30, 165-169. Kimber, I., et al., 1994. Toxicology 93, 13-31. Lalko, J. and Api, A.M. 2006. Food and Chemical Toxicology, 44:739-746. Larsen, W., et al., 2002. Contact Dermatitis, 46(3), 141-144 McNamee, P.M., et al., 2008. Regulatory Toxicology and Pharmacology, 52(1):24-3 Organization for Economic Co-Operation and Development (OECD) 2002. Guideline reference 429. Schnuch, A., et al., 2007. Contact Dermatitis, 57(1), 1-10. Summary and Conclusion • A human dermal sensitization test is not used to determine hazard, rather it is a test to confirm the lack of sensitization at an exposure level which was identified as a NOEL in an animal model or derived as a likely NOEL from quantitative structure-activity relationships. • The induction of dermal sensitization from confirmatory human tests is rare because the assay is used to confirm a NOEL. • The confirmatory HRIPT methodology is robust in design in terms of number of individuals, exposure conditions and evaluation parameters. The test conditions in the HRIPT are exaggerated compared to real life scenarios and are relevant to the generation of data that are very important to the application of a QRA approach. • The EC3 value has recently been demonstrated to closely correlate with the NOEL from confirmatory human sensitization tests designed to confirm lack of induction. • A detailed analysis of the dermal sensitization data in the RIFM database for 57 fragrance ingredients that have exhibited dermal sensitization potential revealed that for the majority of these fragrance ingredients, there is a very good correlation between the predicted NOEL from the murine local lymph node assay and the NOEL in confirmatory human tests. Fragrance Ingredient CAS No. LLNA weighted mean EC3 values (μg/cm2) [no. studies] Potency Human Data WoE NESIL4 (μg/cm2) NOEL HRIPT2 NOEL HMT LOEL 3 Allyl phenoxyacetate 7493-74-5 775 [1]5 Moderate 7094 690 NA 700 α-Amylcinnamaldehyde 122-40-7 2420 [4] Weak 23,6224 NA NA 23,600 α-Amylcinnamyl alcohol 101-85-9 >6250 [1]5 Weak 35434 NA NA 3500 Anisyl Alcohol 105-13-5 1475 [1]5 Moderate NA 34485 NA 1500 Benzaldehyde 100-52-7 > 6250 [1]5 Weak 59