GPO Box 58, Sydney NSW 2001, Australia
Table of Contents
Project Results 3
Individual Project Reports 10
Glyceryl monothioglycolate (GMTG) 10
Cobalt(II) chloride hexahydrate 19
Diazolidinylurea (Germall II) 26
Dowicil 200 36
Imidazolidinylurea (Germall 115) 41
Cl+Me-isothiazolinone (Kathon CG) 50
Abietic acid 57
Zinc dimethyldithiocarbamate (Ziram) 64
Wool alcohols 69
Coconut diethanolamide (Coco. DEA) 83
Basic Red 46 87
Benzalkonium chloride 90
Phenol formaldehyde resin (P-F-R-2) 98
Toluenesulfonamide formaldehyde resin 102
4-tert-Butylphenol, formaldehyde resin (PTBP) 109
Sodium metabisulfite 119
Triethyleneglycol dimethacrylate 125
Appendix - Call for Information 135
Final Report on Hazard Classification of Common Skin Sensitisers
In 2003, the Occupational Dermatology Research and Education Centre (ODREC) noted that many of the top 53 common sensitisers seen in the ODREC clinic and the 361 sensitisers included in commercial European patch testing kits (Chemotechnique Diagnostics) were not classified as sensitisers on the National Occupational Health and Safety Commission’s (NOHSC) ListofDesignatedHazardousSubstances. In response, NICNAS submitted a scoping paper to NOHSC including a screening of a preliminary list of the top 28 ODREC common sensitisers against the NICNAS AustralianInventoryofChemicalSubstances, the NOHSC ListofDesignatedHazardousSubstancesand the EC dangerous substances list – Annex1ofthe28thAdaptationto67/548/EEC. This scoping paper also included results of preliminary literature searches to determine the likely resource requirements for classifying sensitiser chemicals currently missing from the NOHSC list.
On the basis of this initial paper, the NOHSC Chemical Standards Sub Committee at its meeting of 1 July 2003 invited NICNAS to submit a proposal to examine all top 53 ODREC common sensitisers to determine their status with respect to NOHSC hazard classification and, for chemicals currently missing from the NOHSC Hazardous Substances List and the EC Annex 1, to obtain data and classify these against the criteria for sensitisation in the NOHSC ApprovedCriteriafortheClassificationofHazardousSubstances(1999) (NOHSC Approved Criteria).
Of these ODREC common sensitiser chemicals, NICNAS identified a total of 20 individual chemicals that were currently not classified as sensitisers. For these chemicals, NICNAS proposed and in November 2003 NOHSC agreed to NICNAS assessing and classifying these chemicals against the NOHSC Approved Criteria for sensitsation. Specifically, the project proposal included the following actions:
Release of a notice under section 48 of the IndustrialChemicals(NotificationandAssessment)Act1989(Cmwlth) calling for unpublished sensitisation data and any associated adverse incidents during use of these chemicals;
Conduct of literature surveys and compilation of sensitisation data;
Classification of these chemicals against the Approved Criteria for sensitisation and submission of results to NOHSC for adoption to the List of Designated Hazardous Substances.
CorrectionoftheChemicalIdentityofN-CyclohexylbenzothiazylSulphenamideandCl+Me-Isothiazolinone(KathonCG) Further checking of the chemicals to be assessed and classified revealed 2 cases of incorrect identity based on initial information from the Chemotechnique Diagnostics catalogue. For N-Cyclohexylbenzothiazyl sulphenamide, the CAS number was incorrectly listed in the catalogue as 3081-14-9. The correct CAS number was found to be 95-33-0. This chemical is listed in the EC under Annex 1 of the 28thAdaptation to 67/548/EEC.
For the second chemical with the synonyms of Cl+Me-isothiazolinone (Kathon CG), chemical identity information was not available from the catalogue. From an initial search of the CAS National Chemical Inventories (NCI), the chemical was identified as 3(2H)-Isothiazolone, 2-(chloromethyl)-, CAS 21277-94-1. However, a subsequent literature search and information requested from Chemotechnique Diagnostics indicated that this chemical is actually a mixture of two chemicals, 3(2H)- Isothiazolone, 5-chloro-2-methyl- (CAS 26172-55-4) and 3-Isothiazolone, 2-methyl- (CAS 2682-20-4). From a search of the NCI, this mixture itself was found to possess its own chemical identity of 5-Chloro-2-methyl-3(2H)-isothiazolone, mixt. with 2- methyl-3(2H)-isothiazolone, CAS 55965-84-9. Under this chemical identity, this mixture was found to be already listed in the EC under Annex 1 of the 28th Adaptation to 67/548/EEC.
As the scope of this current project only extended to assessing chemicals currently not listed on either the NOHSC ListofDesignatedHazardousSubstancesor the EC dangerous substances list – Annex1ofthe28thAdaptationto67/548/EEC, these two chemicals, N-Cyclohexylbenzothiazyl sulphenamide and Cl+Me-isothiazolinone (Kathon CG), were not assessed as part of this present project.
A notice calling for information was included in the ChemicalGazetteof November, 2003. The notice requested unpublished sensitisation toxicity data and information on any adverse incidents regarding sensitisation by skin contact for 20 sensitiser chemicals. The Gazette Notice is attached (see Appendix).
Written responses were received from two companies. Rohm and Haas Australia Pty Ltd. requested clarification of the scope of assessment for Cl+Me-isothiazolinone and noted concerns with the very low concentration cutoff for Cl+Me-isothiazolinone of 0.0015% to be adopted in the EC. NICNAS clarified with Rohm and Haas that Cl+Me-isothiazolinone (Kathon) was not assessed as part of this project. The assessment of chemicals already listed on either the NOHSC List of Designated Hazardous Substances or the EC Annex 1, as well as the nomination of concentration cutoffs for those chemicals for which hazard classification is appropriate are outside the scope of the present project brief.
Subsequent to the call for information, two of the original 20 chemicals - amerchol and wool alcohols (lanolin) were found not to be separate chemicals. They were thus regarded as the one chemical for assessment and classification.
Data searching was conducted using the US National Library of Medicine database ChemIDplus, the web portal Chemfinder, the International Programme on Chemical Safety INCHEM database, review databases of the American Conference of Governmental Industrial Hygienists (ACGIH), the US Department of Health and Human Services Agency for Toxic Substances and Disease Registry (ATSDR) Toxicological Profiles, online reports of the US National Toxicology Programme (NTP), Cosmetic Ingredient Reviews, Chemical Information System (CIS), the US/Canadian NIOSHTIC with OSHLINE databases, the European Chemicals Review web portal, the Canadian Centre for Occupational Health and Safety (CCOHS) web information service, the Chemical Abstracts Service SCIFINDER system and the bibliographic databases OSHROM, TOMES CPS System, TOXLINE with PUBMED, Current Contents and the Australian Medical Index. The bibliographic databases enabled searching for keywords such as the common names and terms relating to sensitisation in all the major scientific journals including the allergy journal Contact Dermatitis. Lastly, searches were also conducted using the web search engine GOOGLE.
Literature reviews revealed varying amounts and types of information for each of the chemicals. Human data were found for all chemicals in the form of individual case reports showing reactions to particular chemicals. For many, larger surveys or reviews of human patch test results for defined patient populations presenting to dermatology clinics were additionally available. Animal studies were also found for the majority of chemicals. Studies of structure-activity modelling for sensitisation for particular chemicals were also obtained but these were rare.
Animal data were examined with particular emphasis. The NOHSC Approved Criteria notes that hazard classification with the risk phrase R43 is appropriate on the basis of practical experience of skin sensitisation in a substantial number of persons or on the basis of positive animal tests. Regarding animal tests, NOHSC Approved Criteria 4.71 notes that the criteria are met on the basis of animal studies if ≥ 30% and ≥ 15% of animals show positive reactions in adjuvant and non-adjuvant type test methods respectively. OECD Test Guideline 406, adopted in 1992, outlines approved methods for conducting adjuvant and non-adjuvant tests. A more recent OECD Test Guideline 429 outlines sensitisation testing using the local lymph node assay. Therefore, given these specific, quantifiable criteria and the availability of OECD Test Guidelines for both adjuvant and non-adjuvant skin sensitisation tests in animals, the results from such animal studies were given particular scrutiny and priority when classifying against the NOHSC Approved Criteria. Guidance on interpretation of animal studies themselves was obtained from ECETOX (2000) SkinSensitisationTestingforthePurposeofHazardIdentificationandRiskAssessment.MonographNo.29. European Centre for Ecotoxicology and Toxicology of Chemicals, Brussels, Belgium.
In a minority of instances, few or only poor quality animal data were found. In these cases, human patch test datasets and/or individual case reports were the only information available on which the skin sensitisation potential of the chemical could be evaluated. No OECD Guidelines exist for skin sensitisation testing in humans and so in the absence of animal data, each of the human patch test surveys and case reports was evaluated on individual merit. The quality of each study was determined by considering the adequacy of identification of the chemical to be tested, documentation of study methodology and any quality control of patch test readings. In many cases, the evaluation of patch test results were conducted at established dermatological clinics to international standards or guidelines eg. International Contact Dermatitis Research Group (ICDRG) recommendations. Where stated, the institution conducting the study, the standards with which the data were evaluated, the severity of reactions and the prevalence of reaction amongst the test populations were noted.
An important criterion in the NOHSC Approved Criteria for hazard classification on the basis of human data is the prevalence of reactions associated with an individual chemical. This is embodied in Approved Criteria 4.6.6 which states the requirement of “practicalexperienceshowingthatthesubstancesarecapableofinducingsensitisationby skin contact ina substantial number of persons”. Practical experience is also defined as positive data normally in more than one dermatological clinic. For each of the chemicals then, assessment of the sensitisation potential on the basis of human data included not just the consideration of the percentage of the test population showing positive reactions in patch testing, the robustness of the reactions and where they were obtained, but a judgment also as to how representative the test or sample population is to any wider population with potential exposure to the chemical. Even though some studies document a significant percentage of the test population with positive reactions, the representativeness of these test populations to the wider population with potential exposure are sometimes questionable due to admission of patients into the study on the basis of strictly defined, pre-existing allergic or irritant conditions. As expected, few studies were found where the induction of sensitisation with appropriate controls groups was attempted in naïve human subjects. In order then to compare human data to the Approved Criteria, each chemical report in this project considered the prevalence of reactions from human surveys and case reports in the light of potential exposure of the general population to the chemical as estimated from general use information from the surveys themselves or, where noted, from central information sources such as the US National Library of Medicine Hazardous Substances Data Bank (HSDB). In this manner, an assessment was made as to whether the data obtained from defined patient populations represented evidence of skin sensitisation in a substantial number of persons. General guidance on the interpretation of human data was obtained from ECETOX (2002) Use of Human Data inHazardClassificationforIrritationandSensitisation.MonographNo.32. European Centre for Ecotoxicology and Toxicology of Chemicals, Brussels, Belgium.
As well as hazard classifications of individual chemicals, concentration cutoffs for classification of mixtures were also derived. Unfortunately, the majority of the information for sensitisation was sourced from diagnostic human patch tests where, in the cases of positive results, exposure data and, in particular, the concentration of chemical responsible for the initial induction of sensitisation was not known. Also, animal test data where available did not always include concentrations for induction
and challenge. Moreover, even where the common guinea pig maximisation test was available, inferring a concentration cutoff for humans from such a test alone was not possible as the use of adjuvant, intradermal as well as topical injections during induction provides worst-case conditions that precludes direct extrapolation of induction doses to humans. Given these difficulties, no data for any classified chemical suggested deviation from the default concentration cutoff for sensitisation in the NOHSC Approved Criteria of ≥ 1%.
NICNAS also obtained information on cases of allergic reactions as recorded in occupational dermatological databases. Data were obtained from the Occupational Dermatology Research and Education Centre (ODREC) in Melbourne and also from the Centre for Occupational and Environmental Health, University of Manchester. This latter institution maintains two separate occupational health databases, the EPIDERM occupational skin surveillance database and the more general Occupational Physicians Reporting Activity (OPRA) database. For each chemical, the numbers of case entries in each database are provided in the project summaries.
These database entries are a source of information from which the prevalence of reactions to particular chemicals can be inferred. However, the entries themselves cannot be examined and so their relevance to sensitisation potential according to the NOHSC Approved Criteria cannot be confirmed. In this respect, they are presented for information only. They are of only limited value to hazard classification.
Table 1 summarises classification recommendations for 19 assessed chemicals. The common names for the chemicals are those in general use such as in the Chemotechnique Diagnostics Patch Test catalogue.
Hazard classification according to the NOHSC Approved Criteria on the basis of assessment of the single endpoint of skin sensitisation is recommended for 9 chemicals. On the basis of meeting the NOHSC Approved Criteria for skin sensitisation, the risk phrase R43 and hazard category Irritant (Xi) are appropriate for these chemicals.
An additional chemical, cobalt chloride hexahydrate, also meets the NOHSC Approved Criteria for skin sensitisation. However, on the basis of listing of the anhydrous form in the European Union (EU) Annex 1 of the 28thAdaptation to 67/548/EEC, hazard classification on the basis of both skin and respiratory sensitisation is recommended. For this chemical, the risk phrases R42/43 and hazard categories Harmful (Xn) and Irritant (Xi) are appropriate.
As noted previously, two chemicals, N-Cyclohexylbenzothiazyl sulphenamide and Cl+Me-isothiazolinone (Kathon CG) were found already listed in the European Union (EU) Annex 1 of the 28thAdaptation to 67/548/EEC. These were not assessed.
For 7 chemicals, data available for the assessment did not meet the NOHSC Approved Criteria at this time. Lastly, for 2 chemicals, coconut diethanolamide and phenol formaldehyde resin, data were insufficient for classification against the NOHSC Approved Criteria.