Comparison of human skin irritation patch test data with in vitro skin irritation assays and animal data

Transcription

1 Contact Dermatitis 2010: 62: Printed in Singapore. All rights reserved 2010 John Wiley & Sons A/S CONTACT DERMATITIS Comparison of human skin irritation patch test data with in vitro skin irritation assays and animal data Dagmar Jírová 1, David Basketter 2, Manfred Liebsch 3, Hana Bendová 1, Kristina Kejlová 1, Marie Marriott 4 and Helena Kandárová 5 1 National Institute of Public Health, Prague, Czech Republic, 2 DABMEB Consultancy Ltd., Sharnbrook, UK, 3 ZEBET, Federal Institute for Risk Assessment, Berlin, Germany, 4 Safety and Environmental Assurance Center, Unilever Colworth Laboratory, Bedford, UK, 5 MatTek Corporation, MA, Ashland, USA Background: Efforts to replace the rabbit skin irritation test have been underway for many years, encouraged by the EU Cosmetics Directive and REACH. Recently various in vitro tests have been developed, evaluated and validated. Objective: A key difficulty in confirming the validity of in vitro methods is that animal data are scarce and of limited utility for prediction of human effects, which adversely impacts their acceptance. This study examines whether in vivo or in vitro data most accurately predicted human effects. Methods: Using the 4-hr human patch test (HPT) we examined a number of chemicals whose EU classification of skin irritancy is known to be borderline, or where in vitro methods provided conflicting results. Results: Of the 16 chemicals classified as irritants in the rabbit, only five substances were found to be significantly irritating to human skin. Concordance of the rabbit test with the 4-hr HPT was only 56%, whereas concordance of human epidermis models with human data was 76% (EpiDerm) and 70% (EPISKIN). Conclusions: The results confirm observations that rabbits overpredict skin effects in humans. Therefore, when validating in vitro methods, all available information, including human data, should be taken into account before making conclusions about their predictive capacity. Key words: 4-hr human patch test; rabbit test; reconstructed human epidermis model; skin irritation. John Wiley & Sons A/S, Conflicts of interest: Authors have no conflicts of interest. All authors actively and proportionally contributed to the project. Accepted for publication 6 September 2009 The Draize test for skin irritation was developed in the 1940s (1) to predict skin irritation hazard of compounds and formulations that come intentionally or accidentally into contact with human skin. The animal of choice for prediction of the skin corrosion and irritation effects in man was the albino rabbit, due to its high sensitivity to irritating compounds and relatively easy and inexpensive handling. Several modifications of the original Draize procedure were introduced into the test design to improve the predictive potential of this test method. This comprised mainly of shortening the exposure time from 24 hr to 4 hr and abandoning the complementary skin scarification. A possible impact of test animal strain, age, and sex on the test results was also recognized. Despite improvements, the rabbit method has been criticised for low predictivity and questionable relevance towards humans (2, 3), as well as for high subjectivity associated with grading of reactions, the omission of distinct irritation end-points

2 110 JÍROVÁ ETAL. Contact Dermatitis 2010: 62: such as vesiculation, severe eschar formation or ulceration because the grading is based only on erythema and oedema, and finally for low reproducibility (4). Although the animal test proved to be useful in identification of majority of strongly irritating chemicals, still it was failing to correctly predict human non-responsiveness to compounds slightly irritating in rabbits (2, 3). In the 1990s, several large-scale studies on human volunteers were performed that confirmed limitations of the in vivo rabbit test (5 8). These studies concluded that the in vivo rabbit test often overpredicts the severity of skin reactions and damage produced by chemicals, although occasionally, there was also underprediction. Although the rabbit test errs on the side of safety for the consumer, it has been demonstrated that the animal test results should not be relied upon exclusively to identify irritants to human skin (9). In parallel to introducing 4-hr human patch test (HPT) as an ethical tool for assessment of skin effects in man (5 8), significant effort to replace the rabbit skin irritation test by suitable in vitro methods has been exerted. Special emphasis on this subject occurred mainly in Europe under the auspices of ECVAM (10 12). Various in vitro tests have been developed and evaluated, but the best results have been obtained using reconstructed human epidermis (RhE) models (13, 14). Still, one key challenge to be met prior to use of these new methods for regulatory testing purposes is the requirement for their formal validation. Therefore, in 2003, ECVAM initiated a validation trial with the three most promising in vitro tests: SIFT-Skin Integrity Function Test (15), and two RhE-based methods (EpiDerm, EPISKIN) utilizing a common 15 min/42 hr skin irritation test protocol (13, 14). Because the in vivo rabbit skin irritation test is the regulatory accepted method for classification and labelling of chemicals, the Management Team of the ECVAM Skin Irritation Validation Study (SIVS) decided to use results of this test as the in vivo gold standard (16). Although the classification and labelling of chemicals is performed with the view to their application in human practice, existing human data were not taken into consideration when planning SIVS and evaluating the results. Unsurprisingly, during the validation study several false classifications were obtained with both the RhE-based methods, which subsequently decreased their sensitivity and overall accuracy (16). With the aim of understanding the cause(s) of these equivocal results for distinct chemicals, we elected to characterize the skin irritation hazard of these substances in man by means of the 4-hr HPT. The intention was to show the predictive potential of the in vitro methods for species of concern, i.e. the man. Materials and Methods Chemicals The test chemicals are specified in Table 1 regarding their identification, CAS No., purity, and supplier. The chemicals selected for this study were mainly substances whose EU classification is known to be close to the classification cut-off (borderline) and which in prevalidation and validation skin irritation studies (13, 14, 16) gave equivocal results. Before performing HPTs, the general toxicological profile of chemicals was considered; severely corrosive chemicals and chemicals exhibiting other toxicologically unacceptable hazards were excluded. Classifications according to EU and Globally Harmonized System for Classification and Labeling of Chemicals (GHS) rules, which are based on rabbit scores, are documented in Table 2. Besides the classical rabbit scores, Table 2 presents also rabbitdominant median values, calculated according to a new concept developed by Hoffmann et al. (17). The dominant median value for each chemical is determined by calculating the median of the individual rabbit mean scores for each skin reaction (erythema and oedema) and then choosing the larger one, i.e. a dominant one. Information summarized in Table 2 was taken from public sources, such as the European Centre for the Ecotoxicology and Toxicology of Chemicals [ECETOC Database (18)], New Chemicals Database (NCD), which is managed by the European Chemicals Bureau (ECB, a partner unit of ECVAM at the European Commission Joint Research Centre), and TSCA (Toxic Substance Control Act) database maintained by the US Environmental Protection Agency (EPA, Washington, DC, USA), referred by Spielmann et al. (16) and Eskes et al. (19). Two fatty acids (#3 nonanoic and #4 decanoic acids) were included into the dataset as benchmarks for human skin irritation. Chemicals #3, #4, #8, #10, #13, #15, #17, #21 were tested in the past by Basketter et al. (7, 8) and in the current study aimed also to assess transferability of the protocol between laboratories. The 4-hr HPT The selection and recruitment of volunteers and the test methods were carried out in accordance with the ethical principles set out in the Declaration of Helsinki and International Ethical Guidelines for Biomedical Research Involving Human Subjects (20). The 4-hr HPT procedure, performed according to Basketter et al. (8) involves application of chemicals (0.2 ml liquids; 0.2 g moistened solids) on a 25-mm plain Hill Top Chamber containing a Webril

3 Contact Dermatitis 2010: 62: HUMAN SKIN IRRITATION VERSUS ANIMAL/IN VITRO DATA 111 Table 1. Test chemicals No. Chemical identification CAS no. Purity (%) Supplier 1 20% sodium lauryl sulfate r.g. Merck 2 2-Isopropyl-2-isobutyl-1,3-dimethoxypropane Sample provided by CORRELATE 3 Nonanoic acid Aldrich 4 Decanoic acid (capric acid) Aldrich 5 Heptaldehyde (heptanal) Fluka 6 Butyl methacrylate Sigma-Aldrich 7 di-n-propyl disulfide Aldrich 8 Alpha terpineol Alfa Aesar 9 1-Bromohexane >98 Fluka 10 1-Decanol Sigma-Aldrich Undecenoic acid Sigma 12 bis[(1-methylimidazol)-(2-ethyl-hexanoate)], zinc complex Not allocated 95 Sample provided by CORRELATE 13 Hexyl salicylate SAFC 14 Terpinyl acetate SAFC 15 Linalyl acetate >95 Fluka 16 Mixture of isomers: 1-(2-isopropylphenyl)-1-phenylethane, Sample provided by CORRELATE 1-(3-isopropylphenyl)-1-phenylethane, 1-(4isopropylphenyl)-1-phenylethane 17 Heptyl butyrate SAFC 18 Mixture of isomers: 1-(spiro[4.5]dec-7-en-7-yl)pent-4-en-1-one,1-(spiro[4.5]dec Sample provided by CORRELATE en-7-yl)pent-4-en-1-one 19 4-Methylthio-benzaldehyde Aldrich 20 Hydroxycitronellal Aldrich 21 Dodecanoic acid (lauric acid) SAFC 22 di-propylene glycol Aldrich 23 1-Naphthalene acetic acid Sigma 24 3,4-Dimethyl-1-H-pyrazole Sample provided by CORRELATE 25 1-Bromo-4-chlorobutane Aldrich

4 112 JÍROVÁ ETAL. Contact Dermatitis 2010: 62: Table 2. Summary of available in vivo rabbit data No. Chemical identification EU class GHS class Dominating effect No. of tests Rabbit scores dominating effect (average of scores after 24, 48, 72 hr) No. of animals Dominating median (17) Source of in vivo rabbit data and classifications 1 20% sodium lauryl sulfate R38 I Erythema ECETOC, 2 R38 I Erythema n/a n/a 4.0 NCD, Isopropyl-2-isobutyl-1,3- dimethoxypropane 3 Nonanoic acid R34/R38 I Erythema n/a n/a (4.0) Estimated 4 Decanoic acid (capric acid) R38 I Erythema n/a n/a (4.0) Estimated 5 Heptaldehyde (heptanal) R38 I Erythema /3.35* ECETOC, 2 6 Butyl methacrylate R38 I Erythema TSCA, 1 7 di-n-propyl disulfide R38 I Erythema ECETOC, 1 8 Alpha terpineol R38 I Oedema ECETOC, 1, Bromohexane R38 I Erythema ECETOC, 1, Decanol R38 I Erythema ECETOC, Undecenoic acid R38 SLI Erythema ECETOC, 2 12 bis[(1-methylimidazol)-(2- ethyl-hexanoate)], zinc complex R38 SLI Erythema n/a n/a 2.0 NCD, 1 13 Hexyl salicylate R38 SLI Both ECETOC, 1 14 Terpinyl acetate R38 SLI Both ECETOC, 1 15 Linalyl acetate R38 SLI Erythema ECETOC, 2 16 Mixture of isomers: 1-(2-isopropylphenyl)-1- phenylethane,... R38 SLI Erythema n/a n/a 2.0 NCD, 1 17 Heptyl butyrate NI SLI Erythema /1.85* ECETOC, 1 NI NI Erythema n/a n/a 1.3 NCD, 1 18 Mixture of isomers: 1-(spiro[4.5]dec-7-en-7- yl)pent-4-en-1-one, Methylthio-benzaldehyde NI NI Erythema ECETOC, 1 20 Hydroxycitronellal NI NI Erythema ECETOC, 2 21 Dodecanoic acid (lauric acid) NI NI Erythema ECETOC, 2 22 di-propylene glycol NI NI Erythema ECETOC, Naphthalene acetic acid NI NI Both TSCA, ,4-Dimethyl-1-H-pyrazole NI NI Both n/a n/a 0.0 NCD, Bromo-4-chlorobutane NI NI Both ECETOC, 1 1 Spielmann et al. (16), 2 Kandarova et al. (14), *data calculated by Excel spreasheet, ECETOC (18). ECETOC, European Centre for Ecotoxicology and Toxicology of Chemicals; GHS, Globally Harmonized System for Classification and Labeling of Chemicals; n/a, not available; NI, non-irritating to skin; NCD, New Chemical Database [database managed by Europen Chemical Bureau (ECB)]; R38, irritating to skin; R34, corrosive to skin; SLI, slightly irritating to skin; TSCA, Toxic Substances Control Act [(database maintained by the US Environmental Protection Agency (EPA)].

5 Contact Dermatitis 2010: 62: HUMAN SKIN IRRITATION VERSUS ANIMAL/IN VITRO DATA 113 pad (Hill Top Companies, Cincinnati, OH, USA) to the skin of the upper outer arm of 30 human volunteers for up to 4 hr. Exposure time progressively increases from 15 and 30 min through 1, 2, 3 and 4 hr, with each progressive application at a new skin site, until a positive irritation reaction is reported by the volunteer and/or recorded by a trained assessor. Treatment sites are assessed for the presence of irritation using a 4-point scale at 24, 48 and 72 hr after patch test removal. Sodium lauryl sulphate (SLS) at 20% aq. is used as the positive control; water is used as the negative control. A volunteer exhibiting a weak but unequivocal erythema over most or all of the test sites or greater reaction at any of the observation times is considered to have demonstrated a positive irritant reaction and further treatment with that substance does not proceed. The number of panellists who develop a positive irritant reaction after progressive exposure up to 4 hr is determined. The substance is classified as skin irritant (R38), when the incidence of positive irritation reactions to the undiluted test substance is significantly greater than or not significantly different (using Fisher s exact test) than the level of reaction in the same panel of volunteers to 20% SLS. Results The EU and GHS classifications of chemicals based on the available rabbit data, data obtained with the RhE models EPISKIN and EpiDerm, and with the 4-hr HPT are shown in Table 3. According to the EU classification scheme (cutoff 2.0), the dataset of 25 chemicals included 16 irritants and 9 non-irritants. From the 16 tested chemicals reported to be irritant in rabbit, only 5 were found to be irritant for humans. These were SLS (#1), nonanoic acid (#3), decanoic acid (#4), heptaldehyde (#5), and 1-bromohexane (#9). For the positive control (20% SLS), a clear positive skin reaction (erythema) was demonstrated in the whole panel of exposed subjects (94 positive/118 subjects). For the other four irritant substances, a substantial number of positive skin reactions was recorded (nonanoic acid: 19/29; decanoic acid: 28/29; heptaldehyde: 17/29; 1-bromohexane: 16/30). One EU classified irritant, 1-bromohexane (#9), elicited strong sensory irritation in the form of stinging and burning in all volunteers. The exposure time of panellists to 1-bromohexane depended on their subjective perception and ranged from 30 min up to 2 hr. Even the reduced exposure time resulted in clear irritant skin reactions recorded in 16 of 30 volunteers. Sensory irritation was noted to a lesser degree (two subjects) also in case of another EU classified irritant, di-n-propyl disulfide (#7). However, the number of positive skin reactions did not reach the level to qualify the substance as irritant (6/30). All substances classified according to the EU classification scheme as non-irritants proved to be nonirritant in humans. An increased rate of positive skin reactions (11/30) was recorded only for 3,4- dimethyl-1-h-pyrazole; however, the elicited reactions were barely perceptible and borderline to positive, mostly disappearing in 24 hr. The data shown in Table 4 demonstrate the concordance of classifications between the rabbit skin irritation test (EU classification scheme), RhE-based assays and the 4-hr HPT. Results for RhE models were calculated on the basis of outcomes from ECVAM SIVS study (16). When compared with HPT results, the rabbit exhibits 100% sensitivity (five human irritants correctly identified as irritants), but only 45% specificity (only 9 of 20 human non-irritants were correctly identified by the rabbit test as non-irritants), resulting in an overall accuracy of 56%. In our limited dataset, the RhE-based assays exhibited higher accuracy in the prediction of human skin irritation compared with the rabbit test (EPISKIN assay 69.6%, EpiDerm assay 76.0%). The EPISKIN assay provided 100% sensitivity in relation to human data (three out of three human irritants classified correctly) and 50% sensitivity in relation to rabbit data. The specificity values were comparable; 66.7% versus rabbit and 65.0% versus human. The EpiDerm assay turned out to provide sensitivity of 80% when compared with human data (four out of five human irritants classified correctly) and 43% compared with rabbit data. Specificity values were slightly higher than those obtained in EPISKIN assay (77.8% versus rabbit and 75% versus human). Discussion In vitro testing protocols have to be optimized to achieve results comparable or equal to the regulatory accepted reference in vivo methods. However, for many chemicals, including those used in the skin irritation prevalidation, optimization and validation studies (11, 13, 14), in vivo data were relatively scarce, information on interlaboratory reproducibility almost non-existent and reproducibility not systematically assessed. Because in the SIVS several false classifications were obtained with both RhE methods (16), we decided to perform HPTs on these chemicals with no reported toxicological unacceptable hazards in order to define their skin irritation hazard in man. As shown in Table 3, of the 16 tested chemicals classified as irritant in rabbits, only 5 were found to

6 114 JÍROVÁ ETAL. Contact Dermatitis 2010: 62: Table 3. Summary table of in vivo and in vitro results No. Chemical identification EU class (rabbit, cut-off 2.0) GHS class (rabbit) In vivo dominating median 4-hr HPT Positive reactions Positive reactions to SLS Episkin 15 min/42 hr Source of data EpiDerm 15 min/42 hr Source of data 1 20% sodium lauryl sulfate R38 I 4 I 94/118 94/118 R38 2 R38 3 R38 I 4 NI 6/29 26/29 NI-R38-NI 1 NI-NI-NI Isopropyl-2-isobutyl-1,3- dimethoxypropane 3 Nonanoic acid R34/R38 I (4) I 19/29 26/29 Not tested 0 R Decanoic acid (capric acid) R38 I (4) I 28/29 26/29 Not tested 0 R Heptaldehyde (heptanal) R38 I 3.35/4 I 17/29 23/29 R38 2 R Butyl methacrylate R38 I 3 NI 0/30 22/30 R38-R38-R38 1 NI-NI-NI 1 7 di-n-propyl disulfide R38 I 3 NI 6/30 22/30 NI-R38-NI 1 NI-NI-NI 1 8 Alpha terpineol R38 I 2.7 NI 0/29 23/29 R38-R38-R38; (R38) 1; (2) NI-R38-NI; (R38) 1; (3) 9 1-Bromohexane R38 I 2.7 I 16/30 22/30 R38-R38-R38; (R38) 1; (2) NI-NI-NI; (NI) 1; (3) 10 1-Decanol R38 I 2.3 NI 1/30 23/30 R38-R38-R38 1 R38-R38-NI Undecenoic acid R38 SLI 2 NI 1/29 23/29 R38 2 R bis[(1-methylimidazol)-(2- ethyl-hexanoate)], zinc complex R38 SLI 2 NI 0/29 26/29 NI-R38-NI 1 R38-R38-R Hexyl salicylate R38 SLI 2 NI 0/30 22/30 NI-NI-NI 1 NI-NI-NI 1 14 Terpinyl acetate R38 SLI 2 NI 0/30 22/30 NI-R38-NI 1 NI-R38-NI 1 15 Linalyl acetate R38 SLI 2 NI 0/30 23/30 NI 2 NI 3 16 Mixture of isomers: 1-(2-isopropylphenyl)-1- phenylethane R38 SLI 2 NI 0/29 26/29 NI-NI-NI 1 NI-NI-NI 1 17 Heptyl butyrate NI SLI 1.85/1.7 NI 0/30 22/30 NI-NI-NI 1 NI-NI-NI 1 NI NI 1.3 NI 0/29 26/29 NI-R38-NI 1 NI-NI-NI 1 18 Mixture of isomers: 1-(spiro[4.5]dec-7-en-7- yl)pent-4-en-1-one 19 4-Methylthio-benzaldehyde NI NI 1 NI 0/30 22/30 NI-R38-R38 1 NI-NI-NI 1 20 Hydroxycitronellal NI NI 1 NI 0/30 23/30 NI 2 NI 3 21 Dodecanoic acid (lauric acid) NI NI 0.3 NI 0/29 23/29 NI 2 NI 3 22 di-propylene glycol NI NI 0 NI 0/30 22/30 NI-NI-NI; (NI) 1; (2) NI-NI-NI, (NI) 1; (3) 23 1-Naphthalene acetic acid NI NI 0 NI 0/30 22/30 NI-NI-NI 1 NI-NI-NI ,4-Dimethyl-1-H-pyrazole NI NI 0 NI 11/29 26/29 R38-R38-R38 1 R38-R38-R Bromo-4-chlorobutane NI NI 0 NI 0/30 22/30 R38-R38-R38 1 NI-R38-R Spielmann et al. (three laboratories, all runs) (16), 2 Cotovio et al. (one laboratory) (13), 3 Kandárová et al. (one laboratory) (14), 4 Jírová et al. (30). ECETOC, European Centre for Ecotoxicology and Toxicology of Chemicals; GHS, Globally Harmonized System for Classification and Labeling of Chemicals; n/a, not available; NCD, New Chemical Database [database managed by Europen Chemical Bureau (ECB)]; NI, non-irritating to skin; R38, irritating to skin; R34, corrosive to skin; SLI, slightly irritating to skin; TSCA-Toxic Substances Control Act [database maintained by the US Environmental Protection Agency (EPA)].

7 Contact Dermatitis 2010: 62: HUMAN SKIN IRRITATION VERSUS ANIMAL/IN VITRO DATA 115 Table 4. Concordance of classifications Classification methods Sensitivity Specificity Accuracy (%) Rabbit test 4-hr HPT 100.0% (5/5) 45.0% (9/20) hr HPT PT Rabbit test 31.3% (5/16) 100% (9/9) min/42 hr EPISKIN assay Rabbit test 50.0% (7/14) 66.7% (6/9) min/42 hr EpiDerm assay Rabbit test 43.8% (7/16) 77.8% (7/9) min/42 hr EPISKIN assay 4-hr HPT 100.0% (3/3) 65.0% (13/20) min/42 hr EpiDerm assay 4-hr HPT 80.0% (4/5) 75.0% (15/20) 76.0 Concordance of classifications is calculated on the basis of results from ECVAM SIVS study (16). be irritant for humans. This result supports the previously published dataset by Basketter et al. (7, 8), where about 40% of rabbit irritants did not trigger skin irritation when tested by means of 4-hr HPT. In our study, the rabbit test exhibited 100% sensitivity, but only 45% specificity, meaning that the rabbit test identifies all human irritants correctly; however, 55% of non-irritants would be wrongly labelled as irritants (for details see Tables 3 and 4). Notwithstanding a necessary respect for precautionary principles, overlabelling of skin irritation hazard can have a significant economic impact on manufacturing, transport, and end-use of chemicals, e.g. as ingredients in consumer products. The human 4-hr patch test is designed to identify skin irritation hazard, not risk. It is based on the principle of a progressive exposure regime linked to the use of a benchmark positive control, aiming to avoid the production of anything more than mild irritant reactions and meeting the highest ethical standards (21). It has been demonstrated that the test is robust in terms of interlaboratory reproducibility (22) and applicable to all skin types and populations tested (23, 24, 25). Our study results are in agreement with these earlier findings and also confirm the fact that there is no correlation between the susceptibility of an individual to a skin stinging response and subsequent irritation skin reaction development (26, 27). In our group of panellists, the test chemical 1-bromohexane elicited in all volunteers strong sensory irritation in the form of stinging and burning. However, the positive skin reaction was recorded only in 16 of 30 exposed volunteers. In case of di-n-propyl disulfide, two subjects reported sensory irritation but only one developed perceptible skin reaction. This finding supports the generally accepted conclusion that sensory reactions should not be employed for grading (classification) of irritation skin reactions. In addition, this approach reflects the position with animal tests, where sensory effects cannot be measured. While concordance of the rabbit test with results of the 4-hr HPT was rather poor (56%), the RhEbased assays provided more convincing results (EPISKIN assay 69.6% and EpiDerm assay 76.0%). It is probably due to the fact that RhE models mimic the structure of human epidermis better than the animal model. Compared with rabbit skin, RhE models lack hair channels that in occlusive conditions enhance penetration of the chemicals into deep layers of dermis. Additionally, the design of the in vitro tests was adjusted to account for the weaker barrier of RhE models (15 min exposure in RhE models versus 4 hr exposure in human and animal patch tests). In conclusion, 4-hr HPT turned out to be a useful source of information for the weight of evidence approach to reach the correct hazard classification, because it detects acute skin irritation hazard potential using the end-point of concern in the species of concern. It is not intended to predict other types of skin irritant dermatitis, such as cumulative irritant dermatitis (28). Therefore, when approaching validation of alternative methods and developing integrated testing strategies (29), not only the animal data but also appropriate human data should be taken into consideration, as only those provide a final judgement about the predictive ability of new alternative methods. Acknowledgements Authors acknowledge ZEBET at the BfR, Germany, for funding the study, and CORRELATE at JRC, Italy, for providing commercially non-available samples from the ECVAM Skin Irritation Validation Study (16). References 1. Draize J H, Woodward G, Clavery H O. Methods for the study of irritation and toxicity of substances applied topically to the skin and mucous membranes. J Pharmacol Exp Ther 1944: 82: Phillips L, Steinberg M, Maibach H, Akers W. A comparison of rabbit and human skin response to certain irritants. Toxicol Appl Pharmacol 1972: 21: Nixon G A, Tyson C A, Wertz W C. Interspecies comparisons of skin irritancy. Toxicol Appl Pharmacol 1975: 31: Weil C S, Scala R A. Study of intra- and interlaboratory variability in the results of rabbit eye and skin irritation tests. Toxicol Appl Pharmacol 1971: 19: Basketter D A, Chamberlain M, Griffiths H A, Rowson M, Whittle E, York M. The classification of irritants by human patch test. Food Chem Toxicol 1997: 35:

8 116 JÍROVÁ ETAL. Contact Dermatitis 2010: 62: Robinson M K, Whittle E, Basketter D A. A two center study of the development of acute irritation responses to fatty acids. Amer J Contact Dermat 1999: 10: Basketter D A. Contact irritation models. In: Toxicology of Contact Dermatitis: Allergy, Irriancy, and Urticaria, Basketter D A, Gerberick F, Kimber I, Wills C (eds.): Chichester, UK, John Wiley and Sons, 1999: Basketter D A, York M, McFadden J P, Robinson M K. Determination of skin irritation potential in the human 4-h patch test. Contact Dermatitis 2004: 51: Patil S, Patrick E. Animal, human and in vitro test methods for predicting skin irritation. In: Dermatotoxicology Methods: The Laboratory Worker s Vade Mecum, Marzulli F, Maibach H (eds.): Washington DC, Taylor and Francis, 1998: Botham P A, Earl L K, Fentem J H, Roguet R, van de Sandt J J M. Alternative methods for skin irritation testing: the current status. ECVAM Skin Irritation Task Force report 1. Altern Lab Anim 1998: 26: Fentem J H, Briggs D, Chesne C et al. A prevalidation study on in vitro tests for acute skin irritation: results and evaluation by the Management Team. Toxicol in Vitro 2001: 15: Zuang V, Balls M, Botham P et al. Follow-up to the ECVAM prevalidation study on in vitro tests for acute skin irritation. ECVAM Skin Irritation Task Force Report 2. Altern Lab Anim 2002: 30: Cotovio J, Grandidier M H, Portes P, Roguet R, Rubinstenn G. The in vitro acute skin irritation of chemicals: Optimisation of the EPISKIN prediction model within the framework of the ECVAM validation process. Altern Lab Anim 2005: 33: Kandárová H, Liebsch M, Gerner I, Schmidt E, Genschow E, Traue D, Spielmann H. The EpiDerm test protocol for the upcoming ECVAM validation study on in vitro skin irritation tests an assessment of the performance of the optimized test. Altern Lab Anim 2005: 33: Heylings J R, Diot S, Esdaile D J, Fasano W J, Manning L A, Owen H M. A prevalidation study on the in vitro skin irritation function test (SIFT) for prediction of acute skin irritation in vivo: results and evaluation of ECVAM Phase III. Toxicol in Vitro 2003: 17: Spielmann H, Hoffmann S, Liebsch M et al. The ECVAM International Validation Study on In Vitro Tests for Acute Skin Irritation: Report on the validity of the EPISKIN and EpiDerm assays and on the Skin Integrity Function Test. Altern Lab Anim 2007: 35: Hoffmann S, Cole T, Hartung T. Skin irritation: prevalence, variability and regulatory classification of existing in vivo data from industrial chemicals. Regul Toxicol Pharmacol 2005: 41: ECETOC. Skin Irritation and Corrosion: Reference Chemicals Data Bank. ECETOC Technical Report No. 66. Brussels, Belgium, European Centre for Ecotoxicology and Toxicology of Chemicals, Eskes C, Cole T, Hoffmann S, Worth A, Cockshott A, Gerner I, Zuang V. The ECVAM international validation study on in vitro tests for acute skin irritation: selection of test chemicals. Altern Lab Anim 2007: 35: CIOMS. International Ethical Guidelines for Biomedical Research Involving Human Subjects. Geneva, Council for International Organizations of Medical Sciences (CIOMS), Basketter D A, Whittle E, Griffiths H A, York M. The identification and classification of skin irritation hazard by a human patch test. Food Chem Toxicol 1994: 32: Griffiths H A, Wilhelm K-P, Robinson M K, Wang X M, McFadden J, York M, Basketter D A. Interlaboratory evaluation of a human patch test for the identification of skin irritation potential/hazard. Food Chem Toxicol 1997: 35: McFadden J P, Wakelin S H, Basketter D A. Acute irritation thresholds in subjects with Type I Type VI skin. Contact Dermatitis 1998: 38: Basketter D A, Griffiths H A, Wang X M, Wilhelm K-P, McFadden J. Individual, ethnic and seasonal variability in irritant susceptibility of skin: the implications for a predictive human patch test. Contact Dermatitis 1996: 35: Robinson M K, McFadden J P, Basketter D A. Validity and ethics of the human 4-h patch test as an alternative method to assess acute skin irritation potential. Contact Dermatitis 2001: 45: Basketter D A, Griffiths H A. A study of the relationship between susceptibility to skin stinging and skin irritation. Contact Dermatitis 1993: 29: Coverly J, Peters L, Whittle E, Basketter D A. Susceptibility to skin stinging, non-immunologic contact urticaria and acute skin irritation: is there a relationship? Contact Dermatitis 1998: 38: Zhai H, Maibach H I. Dermatotoxicology, 6th edition. Boca Raton, CRC Press LLC, 2004: Hoffmann S, Saliner A G, Patlewicz G, Eskes C, Zuang V, Worth A P. A feasibility study developing an integrated testing strategy assessing skin irritation potential of chemicals. Toxicol Lett 2008: 180: Jírová D, Liebsch M, Basketter D, Spiller E, Kejlová K,Bendová H, Marriot M, Kandárová H. Comparison of human skin irritation and photo-irritation patch test data with cellular in vitro assays and animal in vivo data. Alternat Animal Testing Exp 2008: 14 (Special Issue): Address: Dagmar Jírová, MD, PhD National Institute of Public Health Šrobárova 48, Prague 10, Czech Republic Tel: Fax: jirova@szu.cz