Opinion on Risk of cancer caused by textiles and leather goods coloured with azo-dyes expressed at the 7th CSTEE plenary meeting, Brussels, 18 January 1999.

Assessment of the Report by LGC "The risk of cancer caused by textiles and leather goods coloured with azo dyes"

Opinion

Questions addressed to the Committee:

The CSTEE has been asked to comment on

a) the assessment of the risk of cancer caused by textiles and leather goods coloured with azo dyes as described in the corresponding Report by the Laboratory of the Government Chemist (LGC) and

b) the general quality of the above Report.

Opinion:

The CSTEE considers that the LGC report reviews adequately the situation regarding the risk of cancer for consumers as a result of the use of fabrics dyed with azo compounds and its conclusions are in general acceptable. The Committee supports the Report's recommendation that use of azo dyes should be restricted but a) sees no reason for distinguishing between the 14 amines classified by the EU as category I or II carcinogens and the 8 amines classified by the MAK Commission, and considers that the aim of minimizing or completely avoiding their use should apply equally to all 22 and b) considers that such restriction should not be subject to prior development of validated analytical methodology.

Justification of the Opinion:

The Report reviews the chemistry and toxicology of azo dyes, their use as textile and leather colourants and exposure of consumers using textiles and leather coloured in this way, and discusses the cancer risks associated with such exposure. The CSTEE notes that the Report focuses on consumer risks, even though the wording of the Report's objective ("To consider the risks of cancer caused by textiles and leather goods which are coloured with azo dyes") may be interpreted as also including risks for workers involved in the production and handling of such goods. Exposure of workers involved in the production of such azo-dyed goods is briefly mentioned in the Report, but no discussion of workplace risks is included because it is considered that risks of occupational disease are controlled as a result of best current workplace practices in accordance with health and safety legislation. While this approach is often used in similar assessments, its validity is limited in the absence of information on workplace compliance with current legislation.

The main conclusion of the Report is that, while consumer exposure is likely to be "very low", the associated cancer risks give cause for concern. Based on this conclusion it is recommended by the Report's authors that:

a) use of azo dyes which have the potential to give rise to the 14 aromatic amines classified as Category 1 or 2 carcinogens according to Directive 76/769/EEC should be restricted to the lowest possible levels or completely eliminated,

b) validated analytical methodology should be developed prior to implementation of such restrictions,

c) the socio-economic impacts of any restrictions should be evaluated prior to implementation, and

d) data should be generated to enable the assessment of the mutagenic potential of the remaining 8 MAK III-listed aromatic amines and then of the remaining dyes of interest.

In addition to its main part, the Report includes four Annexes consisting of

I) the conclusions of the "European Inspection of the Notification of New Substances" project which suggest that a significant proportion of new chemicals (including dyes) may in practice be circumventing the notification procedure required by Directive 92/32/EEC, hence not subjected to the corresponding regulatory controls;

II) the chemical formulas of 22 aromatic amines classified as Category 1 or 2 carcinogens according to Directive 76/769/EEC or Category A1 or A2 according to the German MAK Commission;

III) a document titled "Risk assessment for diazo textile dyes", prepared for LGC by Prof. I.C. Shaw; and

IV) a document titled "Best estimate risk assessment calculations for dermal and oral exposure".

In its main part the Report reviews briefly but adequately the background to the use of azo compounds as textile colourants and the relevant legislation (including recent restrictions introduced in some European countries). It points out that the concern from the point of view of carcinogenic risks arises from the potential of azo dyes to undergo in vivo reductive cleavage to aromatic amines, including the 22 amines classified by the EU or the MAK Commission as proven or suspected human carcinogens. Given that the criteria (including the availability of mutagenicity data) for classification in EU Categories 1 or 2 and in MAK Categories A1 or A2 are fundamentally similar, the CSTEE considers that this concern applies equally to azo dyes which can generate the 14 EU-classified amines and to those that can generate the 8 MAK-classified ones and that any restrictions should apply to both groups of azo dyes (it is noted that the amine with CAS No. 120-71-8, classified by the MAK Commission as A2, is 6-methoxy-m-toluidine and not 6-methoxy-o-toluidine as mentioned in the Report).

The discussion of the toxicology of azo dyes is based on the Annex III document. As far as non-cancer effects are concerned, based on detailed studies with 2 azo dyes it is concluded that the NOAEL's for man are likely to be of the order of hundreds of mg/kg and hence unlikely to be of any concern. However, bibliographic evidence is presented suggesting that problems of skin and pulmonary sensitisation, due mainly to azo dyes of the disperse type, may be rarely encountered. Furthermore, it is noted that the animal data on which this conclusion is based refer only to 2 azo compounds, give only a LOAEL and not a NOAEL, and that no allowance for any safety or uncertainty factors are made in the calculations extrapolating to man. However, given that the Report's aim is to assess cancer risks from the use of azo dyes, this omission is understandable.

The mutagenicity and carcinogenicity of azo dyes are discussed in the Annex III report in some detail. No epidemiological observations regarding possible carcinogenic effects of azo dyes in man are available. Despite the fact that only few such compounds have been properly tested for animal carcinogenicity, structure-activity considerations make it likely that many more would prove positive if tested. Reductive cleavage to constituent aromatic amines is suggested as the most important (although perhaps not the only possible) mechanism of genotoxicity. The CSTEE endorses this suggestion and stresses that the conclusions which follow apply only to those azo dyes which do undergo conversion to the specified aromatic amines in vivo.

The main routes of human exposure to azo dyes identified are a) oral ingestion, mainly referring to the sucking of textiles by babies and young children, b) dermal absorption, the route of primary concern for consumers wearing azo compound-dyed products, as well as for workers in dye production and use plants, and c) inhalation, a route of concern for workers in dye production and use industries as well as those handling newly dyed products. Contact with aromatic amines entering the environment through the whole life-cycle of azo dyes in coloured clothes is an additional potential source of human exposure which is not discussed in the Report. A description of human exposure to aromatic amines from sources not related to azo dyes would have been useful but is also not included in the Report.

Turning to the quantitative estimation of consumer exposure resulting from leaching of dyes from fabrics, the only relevant information which appears to be available consists of data on the extractability of azo dyes from fabrics into simulated body fluids and comes from two study reports by the Ecological and Toxicological Association of Dyes and Organic Pigments Manufacturers (ETAD). These data indicate that dye extractability is generally low, with the exception of acid-dyed wool/nylon socks leaching into alkaline perspiration simulant. Taking this as a worst case situation, the total exposure of an adult over a lifetime use of a garment covering the whole body is calculated as 723 µg/kg, while that of a small child which, in addition to suffering similar dermal exposure, receives oral exposure by sucking a piece of garment is calculated at just under 13 mg/kg (a small mathematical error in the Report does not substantially alter the estimated exposure levels). It is concluded that these exposures, which are likely to be gross overestimates, are still far below the human NOAEL for non-cancer end-points. No attempt is made in the Annex III report to obtain a quantitative estimation of the cancer risks. However, the possibility that a child may be exposed to mg quantities of a potentially carcinogenic azo dye is recognised as undesirable by the Report. undesirable by the Report.

Whereas the above calculations were based on a worst case scenario, an attempt is made in Annex IV to derive a series of "best estimates" of exposure to aromatic amines resulting from dyes leaching from fabrics to the skin of adults or children or to the mouth of babies. Employing data on the rate of leaching of various types of dyes taken from the ETAD reports and recommended values of other parameters taken from the Technical Guidance Document for Risk Assessment, and assuming 1% skin penetration, 30% azo compound conversion to amine in the skin (a figure in accordance with the limited experimental evidence) and either 100% or 1% conversion in the mouth, amine exposures in the range 9.7-20 ng/kg/day for adults and 0.48 or 48 ng/kg/day for children are estimated. These exposures are finally used to estimate cancer risks based on the unit risk value for benzidine derived by the US EPA (4.3x10-3 ng/kg/day for 1x10-6 risk). The calculated risks for all cases range from 1 in 100 to 1 in 200,000. While these calculations are likely to err on the high side, and while recognising that some azo dye-related aromatic amines are less powerful carcinogens than benzidine, the magnitude of the estimated cancer risks justifies the conclusion that there may be reasons for concern.

In conclusion, the LGC report reviews adequately the situation regarding the risk of cancer for consumers as a result of the use of fabrics dyed with azo compounds and its conclusions are in general acceptable. However, the Committee sees no reason for distinguishing between azo dyes capable of cleaving to the 14 EU-classified aromatic amines and those capable of cleaving to the 8 MAK-classified ones, and considers that the aim of minimizing or completely avoiding their use should apply equally to azo dyes capable of generating any of these 22 amines.