Isothiazolinones in commercial products at Danish workplaces

Summary: Background: In recent years, a steep increase in the frequency of occupational contact allergy to isothiazolinones has been reported from several European countries. Objective: To examine the extent and occurrence of isothiazolinones in different types of product at Danish workplaces. Methods: Seven different isothiazolinones were identified in the Dictionary ofContactAllergens:ChemicalStructures,Sources, andReferences from Kanerva'sOccupationalDermatitis. By use of the chemical names and Chemical Abstracts Service numbers for these chemicals, information on products registered in the Danish Product Register Database (PROBAS) was obtained. Results: All seven isothiazolinones were registered in PROBAS. The top three isothiazolinones registered were: benzisothiazolinone (BIT), registered in 985 products, methylisothiazolinone (MI), registered in 884 products, and methylchloroisothiazolinone (MCI)/MI, registered in 611 products. The concentration ranges were 0.01 ppm to 45% for BIT, 0.01 ppm to 10% for MI, and 0.01 ppm to 14.1% for MCI/MI. The most common product type was ‘paint and varnish’; five of the seven isothiazolinones were registered in this type of product. Conclusion: Isothiazolinones are present in multiple products registered for use at workplaces, and may occur in high concentrations.

2 ‐ methyl ‐ 4,5 ‐ trimethylene ‐ 4 ‐ isothiazolin ‐ 3 ‐ one; allergic contact dermatitis; benzisothiazolinone; dichlorooctylisothiazolinone; isothiazolinones; methylchloroisothiazolinone; methylisothiazolinone; octylisothiazolinone

Knowledge regarding the individual and general exposures to hazardous chemicals in our environment is pivotal for understanding the individual and general disease risk related to these chemicals. Such information is also essential for the planning of preventive initiatives. Information on the production and use of the chemicals may be retrieved from national and international statistical offices, databases, and manufacturers; this data collection is difficult and time ‐ consuming. The Scandinavian countries have developed registers that contain information on the contents of hazardous chemicals in products registered for occupational use in the respective countries.

The isothiazolinones are preservatives (Table [NaN] ) that have been in use for > 30 years in products for both occupational and consumer use [1] . This group of chemicals was soon recognized as containing strong or extremely potent allergens [2] . Nevertheless, these chemicals were permitted (Table [NaN] ), and registered in different product categories [3] . The general argument made in risk assessment has been that even extremely potent contact allergens can be used safely in products if the exposure concentration is sufficiently low [4] . The risk assessment methods rest primarily on animal assays such as the guinea ‐ pig maximization test (GPMT) [5] or the local lymph node assay (LLNA) [6] ,[7] , and/or experimental induction studies in humans (human repeated insult patch test) [8] . However, these methods have either failed or been incorrectly interpreted with regard to fulfilling the expectations of safety in terms of the risks of sensitization [1] ,[6] ,[7] ,[9] ,[10] . The consequences have been serious, and have given rise to numerous occupational cases of severe allergic contact dermatitis [11] ,[12] ,[13] ,[14] ,[15] ,[16] ,[17] and an epidemic of contact allergy resulting from the use of cosmetic products caused by methylisothiazolinone (MI) and methylchloroisothiazolinone (MCI) [1] ,[18] ,[19] ,[20] ,[21] .

Marketed isothaizolinones and their regulation

1 BPR, biocidal products regulation; CLP, the regulation on classification, labelling and packaging of substances and mixtures; INCI, International Nomenclature of Cosmetic Ingredients; PT, product type; REACH, Registration, Evaluation, Authorization and Restriction of Chemical substances; RoI, registry of intention – ‘warning’ of future proposal.

• 2 * According to CLP there is non ‐ specific concentration limit for MI but according to the inventory there is a “self ‐ classification” with a specific concentration limit on 0.1%. This concentration is not validated or authorized but is used by some manufactures because of the allergenic potency of MI. For those manufacturer this concentration will be effective from 1st of June 2015.
• 3 ** The European Commission requested Slovenia to classify MI according to the Annex XV dossier in REACH. The deadline was December 2013.

In view of the above, we provide an overview of the occurrence of isothiazolinones in registered chemical products, and discuss the clinical and general health implications.

We selected the isothiazolinones listed in the Dictionary of Contact Allergens: Chemical Structures, Sources, and References from Kanerva's Occupational Dermatitis[33] .

In the examination of product types registered with content of the isothiazolinones, the chemical names and the Chemical Abstract Service (CAS) numbers for these isothiazolinones were searched for in the Danish Product Register Database (PROBAS) (September 2012).

PROBAS is a database in which the composition of, primarily, hazardous chemical products for occupational use is registered. Products are registered: (i) if the product/substance is manufactured or imported for occupational use in a quantity of > 100 kg annually; (ii) if the product contains at least one chemical that is registered as harmful according to the Danish Ministry of the Environment and the Danish Working Environment Authority (WEA); (iii) if the product contains ≥ 1% of the substance (for preservatives, the limit is 0.1%); (iv) if the product/substance is assigned an occupational exposure limit in the WEA list of limit values for substances and materials; and/or (v) if materials contain ≥ 1% of a substance that has been assigned an occupational exposure limit in the WEA list of limit values for substances and materials [34] . When a product is registered in PROBAS, we assume that it is used in the Danish work environment.

PROBAS is updated at the end of every odd year with data collected from the manufacturers in even years. In cases where there were fewer than three manufacturers, the specific product types were classified as confidential.

In PROBAS, the chemicals are categorized according to the Use Categories Nordic (UCN) code system [35] . The system is the same in Norway, Sweden, and Denmark. The system consists of main groups and subgroups. The code for the main group consists of three characters, and the code for the subgroup consists of the main group plus three digits. One chemical can have more than one UCN code. According to the principal manufacturer of MCI/MI, MCI is not sold as an independent substance, so, even though it is listed on its own in PROBAS, it is likely to be used together with MI.

All confidential information was omitted from the dataset. No main groups were deleted, but subgroups that had a ratio of < 2% between the main group and the subgroup were deleted by the authors to maintain a more relevant overview.

The isothiazolinones included in this study are shown in Table [NaN] . All seven isothiazolinones were registered in PROBAS. The results of the search in the database for different types of registered product are shown in Tables [NaN] and [NaN] . The concentrations listed in the two tables are in the same format as used in PROBAS. It is not possible to compare the different product types, because some of them are raw materials and others are products for the downstream user.

The results from the Danish Product Register Database (PROBAS) for benzisothiazolinone, methylisothiazolinone, methylchloroisothiazolinone, and Methylchloroisothiazolinone/ methylisothiazolinone

4 * Percentages.

The results from the D anish P roduct R egister D atabase (PROBAS) for dichlorooctylisothiazolinone and octylisothiazolinone

5 * Percentages.

BIT was registered in 985 different products registered in PROBAS, and was the isothiazolinone most often found in products registered in PROBAS. The top three product types containing BIT were paints and varnishes (544), cleaning/washing agents (108), and polishing agents (65). BIT was registered in concentrations from 0.01 ppm to 45%.

MI was the second most frequently registered isothiazolinone, with 884 different products registered in PROBAS. The top three product types containing MI were paints and varnishes (471), cleaning/washing agents (87), and polishing agents (60). MI was registered in concentrations from 0.01 ppm to 10%.

MCI/MI was the third most frequently registered isothiazolinone, with 611 different products registered in PROBAS. The top three product types containing MCI/MI were paints and varnishes (363), cleaning/washing agents (50), and polishing agents (32). MCI/MI was registered in concentrations from 0.01 ppm to 14%.

MCI was the fourth most frequently registered isothiazolinone, with 474 different products registered in PROBAS. The top three product types containing MCI were paints and varnishes (275), cleaning/washing agents (34), and biocides (25). MCI was registered in concentrations from 0.01 ppm to 17%.

OIT was the fifth most frequently registered isothiazolinone, with 111 different products registered in PROBAS. The top three product types containing OIT were paints and varnishes (60), biocides (16), and filling agents (5). OIT was registered in concentrations from 0.02 ppm to 16.3%.

DCOIT was registered in 58 different products, which was the second lowest number among the isothiazolinones in this study. The top three product types containing DCOIT were paints and varnishes (38), biocides (12), and filling agents (5). DCOIT was registered in concentrations from 94 ppm to 28.1%.

MTMIT was the only isothiazolinone for which all of the use categories (UCN codes) were confidential. The chemical was registered in seven products in concentrations from 47.6 to 150 ppm.

The present study shows that thousands of products are registered in PROBAS as containing isothiazolinones.

BIT and MI were registered in 985 and 884 products, respectively; MCI/MI was registered in 611 products.

The most prominent product type containing these three isothiazolinones was paints and varnishes (with a total of 3567 products), in which BIT, MI and MCI/MI were the predominant isothiazolinones. This is in accordance with clinical experience (Table [NaN] ), where isothiazolinone allergies have been reported in painters [36] ,[37] ,[38] ,[39] and paint production workers [12] ,[40] ,[41] . It is also known that MCI/MI and MI may evaporate and cause airborne allergic contact dermatitis, a manifestation that has been increasingly seen with particular MI ‐ preserved paints [36] ,[37] ,[38] ,[42] . MI and BIT have been found to cause airborne contact dermatitis, respiratory symptoms, including acute asthma, and systemic allergic contact dermatitis [15] ,[36] . The presence of MI in paints was further highlighted in a study in which, of 19 different water ‐ based paints from the Danish retail marked, all contained MI, four contained MCI, and 16 contained BIT [43] . The emission of BIT, MCI and MI from paint has also been shown [43] .

Clinical effects of cutaneous and airborne exposures to the isothiazolinone preservatives

6 BIT, benzisothiazolinone; MCI, methylchloroisothiazolinone; MI, methylisothiazolinone.

Preservative exposure from paints has changed since 2000, when MI was introduced as a stand ‐ alone preservative and, to some extent, replaced MCI/MI. The reason may be that, whereas MCI/MI is on the list of dangerous substances, labelled R43/H317 (may cause sensitization), and has to be declared on the label if it is present in amounts > 15 ppm, no such requirements exist for MI (Table [NaN] ); instead, there is self ‐ classification, with a specific concentration limit of 0.1% in industrial products [25] . This may erroneously lead the manufacturers to assume that MI is unproblematic [29] .

The products registered with the highest content of isothiazolinones were car care products, BIT being present in 21.9% of products, and MI in 18.8%; a similar pattern was seen for polishing agents for cars and floor products (Tables [NaN] and [NaN] ). We have not found any reports concerning contact allergy related to the use of such products.

Contact allergy to isothiazolinones in the paper and textile industry has also been reported [14] ,[17] ,[48] ,[49] . In our study, BIT, MI and MCI/MI were registered in products for the surface treatment of paper, cardboard, and other non ‐ metals, although not very frequently. In the paper industry, they may be present in the concentrated biocides that are used and added to the pulp. All of the isothiazolinones were registered as biocide products. This is a product group that may have multiple uses.

Contact allergy to BIT and OIT has been reported in metalworkers [50] ; these are also the two allergens found in cooling agents by chemical analysis in a Finnish study [51] . In PROBAS, only OIT was registered in cooling agents for metal processing (Table [NaN] ). This emphasizes the importance of patch testing with these special isothiazolinones in metalworkers with occupational eczema.

Exposure to high concentrations of isothiazolinones (Table [NaN] ) can cause chemical burns followed by contact sensitization resulting from a single exposure [52] ,[53] ,[54] ,[55] ,[56] ,[57] . Further accidental exposure to such high concentrations may lead to generalized dermatitis accompanied by systemic contact dermatitis and subjective symptoms [15] . Such cases may develop into chronic diseases.

A part of the problem is that much emphasis and reliance has been put on the results of animal assays, in particular the LLNA, to predict the sensitization potency of these substances. The reporting of these results is very brief in the scientific literature, as EC3 values only, not supported by data, and has also been shown to be misleading [9] . This is in contrast to the amount of human data from the occupational setting and consumers, which, for a long period of time, have been ignored by industry and regulators.

In three product categories, namely paints and varnishes, biocides, and filling agents, all six isothiazolinones were registered, whereas, for all other product categories, some but not all isothiazolinones were used.

There are only few studies on cross ‐ reactions between the different isothiazolinones. In 1992, Damstra et al. found, in an analysis of 556 patch tested patients, that 8 of 10 patients who reacted to BIT also reacted to MCI/MI, and 8 of 56 patients who reacted to MCI/MI also reacted to BIT [58] . In a recent study, Mose et al. found that all patients with concomitant positive patch test reactions to OIT, BIT and MCI/MI were painters, which suggests concomitant sensitization rather than cross ‐ reactivity [16] . Cross ‐ reactivity can only be studied in animals or if the primary sensitizer in humans is known. In GPMTs, it was shown that cross ‐ reactivity between MCI and MI exists, whereas no cross ‐ reactivity could be shown between MCI, MI as primary sensitizer, and OIT, and probably BIT [59] . In 2008, Isaksson et al. also showed, by testing workers accidentally sensitized to MCI, that cross ‐ reactivity between MCI and MI is likely, depending on the degree of contact allergy [60] , which is in accordance with clinical experience [61] . This means that, in detecting cases of contact allergy to isothiazolinones, one cannot rely on testing only with MCI and MI in the baseline series.

It is very important for the physician and the clinical procedure that all information concerning the product be on the product label and/or in the Material Safety Data Sheet (MSDS) or other product datasheets, so that exposures to allergens can be identified, appropriate patch testing performed, and the correct diagnosis made. Otherwise, important allergens can be missed. The correct diagnosis is very important for patients with occupational allergic contact dermatitis, both for the medical prognosis and for the compensation, depending on national laws [62] .

In conclusion, isothiazolinones are present in multiple products registered for use at the workplace, and may occur in high concentrations. BIT was the most frequent isothiazolinone, being registered in the most products (985), and the concentration range was 0.01 ppm to 45%.

Knowledge of the use and exposure to the isothiazolinone preservatives, together with the contact allergic potential and the clinical disease related to the exposure, leads to the following points to be addressed:

The use concentrations in the different exposure scenarios need to be revised.

All types of isothiazolinone, independently of concentration, should be labelled on the product and included in the MSDS.

The use of concentrated solutions of isothiazolinones should only be permitted in closed systems, and the workers involved need proper education.

Is the margin of safety for isothiazolinones so low that future use should be abolished?

We thank the Danish Working Environment Research Fund for funding this project, and give special thanks to Peter Herskind (chief consultant in the work environment at the Confederation of Danish Industry) and Jan Toft Rasmussen (consultant at the Danish Metalworkers' Union) for participating in the steering group and for valuable help throughout the project, to Lea Stine Tobiassen (toxicologist at the Danish Environmental Protection Agency) for valuable help throughout the project, and to Poul Erik Andersen at the Danish Product Register for performing the analysis in PROBAS.