1.1.1. Nomenclature

• Chem. Abstr. Serv. Reg. No.: 28407-37-6 (ECHA, 2022)
• Chem. Abstr. Serv. name: cuprate(4-), [µ-[[3,3'-[[3,3'-di(hydroxy-κO)[1,1'-biphenyl]-4,4'-diyl]bis(2,1-diazenediyl-κN1)]bis[5-amino-4-(hydroxy-κO)-2,7-naphthalenedisulfonato]](8-)]]di-, sodium (1:4)
• EC No.: 249-008-8 (ECHA, 2022)
• IUPAC systematic name: tetrasodium;5-amino-3-[[4-[4-[(8-amino-1-hydroxy-3,6-disulfonatonaphthalen-2-yl)diazenyl]-3-hydroxyphenyl]-2-hydroxyphenyl]diazenyl]-4-hydroxynaphthalene-2,7-disulfonate; copper (NCBI, 2020)
• Synonyms: DIRECT BLUE 218, 28407-37-6; CI 24 401; Fastusol Blue 9GLP; Solantine Blue 10GL; Pontamine Bond Blue B; Amanil Supra Blue 9GL; CI Direct Blue 218; Pontamine Fast Blue 7GLN; UNII-RR3V5FL20N; RR3V5FL20N; CCRIS 6142; HSDB 4223; NCI C60 877; EINECS 249-008-8; INTRALITE Blue 8GLL; DIRECTBLUE218; (3,3′-((3,3′-Dihydroxy-1,1′-biphenyl-4,4′-diyl)bis(azo)bis(5-amino-2,7-naphthalenedisulfonato-(O4,O3)))dicopper, tetrasodium salt; 2,7-Naphthalenedisulfonic acid, 3,3′-((3,3′-dihydroxy(1,1′-biphenyl)-4,4′-diyl)bis(azo)bis(5-amino-4-hydroxy-, sodium salt, copper complex; Copper, [mu-[tetrahydrogen-3,3′-[(3,3′-dihydroxy-4,4′-biphenylylene)bis(azo)]bis[5-amino-4-hydroxy-2,7-naphthalenedisulfonato](4-)]di-, tetrasodium salt; Cuprate(4-), [mu-[[3,3′-[(3,3′-dihydroxy[1,1′-biphenyl]-4,4′-diyl)bis(azo)]bis[5-amino-4-hydroxy-2,7-naphthalenedisulfonato]](8-)]]di-, tetrasodium; 2,7-Naphthalenedisulfonic acid, 3,3′-[(3,3′-dihydroxy[1,1′-biphenyl]-4,4′-diyl)bis(azo)]bis[5-amino-4-hydroxy-, sodium salt, copper complex; 1-Naphthol-3,6-disulfonic acid, 2,2′-(3,3-dihydroxy-4,4′-dipiphenylylenebisazo)bis[8-amino-, dicopper deriv., tetrasodium salt (CAMEO, 2020; NCBI, 2020).

1.1.2. Structural and molecular formulae, and relative molecular mass

• 
• Molecular formula: C₃₂H₂₀Cu₂N₆Na₄O₁₆S₄ (PubChem CID: 34237 or 137241407)
• Relative molecular mass: 1091.9 (powder form)
• A solution form of this dye is indicated as PubChem CID: 24832073, with the molecular formula C₃₂H₁₆Cu₂N₆Na₄O₁₆S₄ and a relative molecular mass of 1087.8.
• [The Working Group noted that the chemical and physical properties presented below are for the powder form.]

1.1.3. Chemical and physical properties of the pure substance

• Description: deep purple to dark blue amorphous powder (NTP, 1992; ECHA, 2020a)
• Boiling point: 1560 °C (estimated, based on EPI Suite MPVPBP V1.43) (ECHA, 2020a)
• Melting point: 350 °C (estimated, based on EPI Suite MPVPBP V1.43) (ECHA, 2020a)
• Density: 2.8 ± 0.1 g/cm³ at 20 °C (ECHA, 2020a)
• Solubility: 1.00–5.00 × 10⁻⁴ mg/L at 17 °C in water (ECHA, 2020a)
• Vapour pressure: 1.06 × 10⁻³⁹ Pa at 25 °C (ECHA, 2020a)
• Flash point: flash-point data for this chemical are not available; however, it is probably combustible (NTP, 1992)
• Stability and reactivity: CI Direct Blue 218 is a diazo compound. Azo, diazo, and azido compounds can detonate. Their nitro groups facilitate rapid decomposition. This particularly applies to organic azides that have been sensitized by the addition of metal salts or strong acids. Toxic gases are formed by mixing materials of this class with acids, aldehydes, amides, carbamates, cyanides, inorganic fluorides, halogenated organics, isocyanates, ketones, metals, nitrides, peroxides, phenols, epoxides, acyl halides, and strong oxidizing or reducing agents. Flammable gases are formed by mixing materials in this group with alkali metals. Explosive combination can occur with strong oxidizing agents, metal salts, peroxides, and sulfides (CAMEO, 2020).
• Octanol/water partition coefficient (P): log K_ow, −0.77 (ECHA, 2020a).

1.1.4. Impurities

The purity of CI Direct Blue 218 was reported to be approximately 60% in a study by the National Toxicology Program (NTP) (NTP, 1994). Chemical characterization of two lots characteristic of the product used by industry indicated more than 12 impurities, of which the majority appeared to contain the reducible azo bond (Morgan et al., 1994; NTP, 1994). No attempt was made to identify the chromatographic peaks; however, reduction titration of azo groups indicated that the two lots had purities of 90% and 83%. The concentrations of benzidine and 3,3′-dimethoxybenzidine were determined. Benzidine could not be detected in either lot at levels greater than 1 ppm [1 µg/mL]. 3,3′-Dimethoxybenzidine was found at levels less than or equal to 7 ppm [7 µg/mL], but the level was less than 1 ppm in the lot used for a 2-year bioassay (Morgan et al., 1994; NTP, 1994). The study by the NTP also noted that other impurities, in addition to those containing azo groups, probably included inorganic copper salts.

[The Working Group noted that the only data on impurities were derived from the abovementioned assays reported by the NTP in 1994. However, several manufacturers listing CI Direct Blue products online in 2021 were claiming a purity of 96–99%.]

1.2.1. Production process

CI Direct Blue 218 is produced by coupling one mole of ortho-dianisidine (3,3′-dimethoxybenzidine) to two moles of 4-amino-5-hydroxy-2,7-naphthalene disulfonic acid under alkaline pH conditions (resulting in CI Direct Blue 15), followed by the addition of a copper salt and the elimination of methyl groups from the methoxides to form a copper complex (Kirk-Othmer, 1978). Due to the copper chelation process, the substance does not contain methoxy groups that are characteristic of the ortho-dianisidine (3,3′-dimethoxybenzidine) moiety, which is a major component of the dye (NIOSH, 1980).

1.2.2. Production volume

CI Direct Blue 218 is listed by the Organisation for Economic Co-operation and Development (for the year 2007) as a High Production Volume chemical (OECD, 2009). The production volumes in the USA were 4.03 × 10⁵ kg in 1977, 3.54 × 10⁵ kg in 1979, and 3.33 × 10⁵ kg in 1985 (NCBI, 2003). In 1994, there were four production plants registered in the USA (NCBI, 2003).

Production or import volumes in the USA were reported to be between 10 000 and 500 000 pounds [between 4.5 and 230 tonnes] in 1986, 1990, 1994, 1998, and 2002, and < 500 000 pounds [< 230 tonnes] in 2006 (US EPA, 2003, 2007; IARC, 2010). The national aggregated production volumes in the USA according to Chemical Data Reporting records were < 1 000 000 pounds [< 450 tonnes] per year from 2012 to 2015 (US EPA, 2020). In 2020, CI Direct Blue 218 was available from two suppliers in the USA and 12 suppliers from China (Chemical Register, 2020). Based on Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) registration data reported in 2018, between 1 and 10 tonnes of CI Direct Blue 218 were manufactured or imported in the European Economic Area (ECHA, 2020a).

[The Working Group noted that no specific information could be found on production volumes of CI Direct Blue 218 in China or elsewhere outside of the USA.]

1.2.3. Use

CI Direct Blue 218 is used as a dye for cellulose, acetate, nylon, silk, wool, tissue, fine papers, and textile goods with a urea-formaldehyde finish (NCBI, 2003; ECHA, 2020a). As noted in the introduction to this section, direct dyes have a natural affinity for cellulose without the need for a mordant (Waring & Hallas, 1990; Aspland, 1991). Azo dyes are synthetic compounds with vivid colours, of which > 7 × 10⁵ tons [> 6.4 × 10⁵ tonnes] are produced annually worldwide, accounting for > 50% of all dyestuffs produced worldwide (Chung, 2016). Boeniger (1980) reported the total use of benzidine-based dyes as follows: 40% to colour paper, 25% to colour textiles, 15% to colour leather, and 20% in diverse applications in the petroleum, rubber, plastics, wood, soap, fur, and hair dye industries (Boeniger, 1980). [The Working Group noted that these data were not specific to CI Direct Blue 218, for which no data were available.]

1.3.1. Air

No specific methods are available for the measurement of CI Direct Blue 218 in air. National Institute for Occupational Safety and Health (NIOSH) method No. 5013 is available for benzidine-based dyes (NIOSH, 1994). This method also can be used for ortho-dianisidine-based (3,3′-dimethoxybenzidine-based) dyes. Sampling of airborne dyes is performed with a 5 µm polytetrafluoroethylene membrane filter with a flow rate of 1–3 L/minute. After ultrasonic treatment of the filter, sodium hydrosulfite is added for reductive cleavage of the dye; the isolated benzidine is finally analysed by high-performance liquid chromatography-ultraviolet (HPLC-UV). The estimated limit of detection (LOD) of the method is 6–20 µg/m³.

1.3.2. Other environmental media

Few specific methods are available for the measurement of CI Direct Blue 218 and dimethoxybenzidine-based dyes in environmental media. A method for the chemical characterization of CI Direct Blue 218 was used that applied thin-layer chromatography and HPLC with UV/visible light detection at 254 and 658 nm (NTP, 1994). CI Direct Blue 218 was also measured in feed following extraction with methanol or methanol with tetrabutylammonium hydroxide by spectrophotometric measurement of the 622 nm absorbance maximum. Different analytical methods are used for the analysis of 3,3′-dimethoxybenzidine and dimethoxybenzidine-based dyes in a variety of matrices (water, paint, textiles, food, and toys). In general, methods involve reductive cleavage of the dye to the free amine and analysis of the resultant amines. [The Working Group noted that, in the case of CI Direct Blue, it is unlikely that 3,3′-dimethoxybenzidine is formed after cleavage, since the methoxy groups are no longer present. Instead 3,3′-dihydroxybenzidine is more likely to be formed. The analytical methods described below are also relevant for these products because they detect the amine groups.]

While gas chromatography (GC) analysis invariably requires derivatization of the amine before the analysis, analysis by liquid chromatography (LC) in combination with mass spectrometry (MS) does not. Also, the use of modern liquid chromatography with tandem mass spectrometry (LC-MS/MS) methods permits the analysis of complex mixtures. A method based on the analysis of amine with HPLC-UV (after reductive cleavage) has been developed for 3,3′-dimethoxybenzidine-based dyes in toys, with an estimated LOD of 0.2 µg/g (Garrigós et al., 2002). For wastewater, a method has been developed by dissolving the dyes in methanol, dichloromethane, or ethylacetate before analysis with GC-MS (Doherty, 2005). For textiles, a method that involves refluxing with chlorobenzene followed by three consecutive extractions with citrate buffer, hydrosulfite, and tert-methyl ether was developed for use with LC-MS/MS analysis (Sutthivaiyakit et al., 2005). This method forms the basis for the later-developed International Organization for Standardization (ISO) method 14362-1:2017 (ISO, 2017). In the ISO method, several chromatographic techniques and detection methods are described: HPLC with diode-array detection, HPLC-MS, GC-MS, and GC-flame ionization detection. The estimated LOD of the method is 5 µg/g.

1.3.3. Biological specimens

No biomonitoring methods for the assessment of exposure to CI Direct Blue 218 or its possible metabolites were identified in the literature.

1.4.1. Environmental occurrence

A study of the Yamaska River in Quebec, Canada, conducted between 1985 and 1987, investigating the occurrence of 15 dyes in river water, sediments, and solids, as well as fish downstream of textile mills, did not detect CI Direct Blue 218 (Maguire, 1992). No other environmental studies on CI Direct Blue 218 were identified. Very little is known about the environmental occurrence, persistence, and fate of individual dyes because of difficulties in determining different chemical classes of dyes at trace levels in environmental samples (Maguire, 1992). Benzidine and its congeners are not known to occur naturally in the environment. Although few actual measurements of the release of benzidine-based compounds into the environment have been reported, it is thought that manufacturing and processing plants for dyes and pigments derived from benzidine and its congeners are the major sources of release. Three major sources of environmental release of dyes and pigments derived from benzidine and its congeners have been identified: process wastewaters; atmospheric release; and disposal of dyed articles (US EPA, 1980).

About 10% of the azo dyes used in textile dyeing processes are released into the environment (Chung, 2016; dos Santos, 2018). In the 1980s, it was estimated that, worldwide, 280 000 tons [units assumed to be imperial tons; 284 494 metric tonnes] of textile dyes were annually discharged into industrial effluents (Chung, 2016). Since the azo dyes represent about 70% by weight of the dyestuffs used, it follows that they are the most common group of synthetic colourants released into the environment (Chung, 2016).

Generally, ionic azo dyes released into surface waters or wastewater are expected to bind primarily to suspended organic matter, due to electrostatic interactions, and ultimately sequester to sediments or wastewater sludge. However, a proportion of ionic dyes are likely to remain dissolved in the water column (dos Santos, 2018). Because of their stability and microbial resistance, azo dyes are not readily removed from wastewater by conventional treatment methods (Chung, 2016). Contaminated sewage sludge is often dumped in landfill waste sites, resulting in soil and groundwater contamination (Tkaczyk et al., 2020). Furthermore, sludge from wastewater treatment plants or industrial wastewater may be applied to agricultural fields, which may result in significant concentrations of dyes in agricultural soils (dos Santos, 2018; Tkaczyk et al., 2020). Complexed metal dyes have half-lives of 2–13 years and in the aquatic environment the heavy metal cations can be assimilated by fish gills, which can lead to accumulation in certain tissues (Lellis et al., 2019). As a consequence of both the direct and indirect sources described above, synthetic organic dyes pass through different trophic levels of the food web, from (water) plants and algae through consumers of the first order (e.g. crustaceans) and through secondary consumers (e.g. fish) to humans. Since this may cause biomagnification of dye components in the food chain, these compounds are considered to be persistent bioaccumulative toxic substances (Tkaczyk et al., 2020).

1.4.2. Occupational exposure

From a survey conducted between 1981 and 1983, NIOSH estimated that a total of 12 290 workers might come into contact with CI Direct Blue 218 in the textile and paper industries (NIOSH, 2017). Industrial exposure to dyes may occur through inhalation of dust or mist, accidental ingestion, or direct contact with the skin (NIOSH, 1980). Potential exposure to CI Direct Blue 218 occurs during the manufacturing process (synthesis, processing, packaging, transportation, or maintenance and clean-up), from the application of the dye on products, and from further processing of dyed products that results in particles being formed (NIOSH, 1983).

1.4.3. Exposure in the general population

No data on exposure to CI Direct Blue 218 in the general population have been identified; however, most environmental exposure to 3,3′-dimethoxybenzidine and 3,3′-dimethoxybenzidine-based dyes has been described as resulting from contact with contaminated air, water, or soil. In addition, the general population may also be exposed to 3,3′-dimethoxybenzidine-based dyes via contact with paper or fabric products containing these dyes, or through consumer use of the dyes (NTP, 2016). However, the benzidine-based dyes in finished products are not considered to migrate from the product as a result of washing, perspiration, or contact with saliva (US EPA, 1980).

Oral administration (feed)

In a study of chronic toxicity and carcinogenicity that complied with Good Laboratory Practice (GLP) and that was conducted by the NTP (1994), groups of 50–51 male and 50 female B6C3F₁ mice (age, 7 weeks) were given feed containing CI Direct Blue 218 (a desalted commercial dye of 60% copper complex of 3,3′-[(3,3′-dihydroxy[1,1′-biphenyl]-4,4′-diyl)bis(azo)]bis[5-amino-4-hydroxy-2,7-naphthalenedisulfonic acid] tetrasodium salt; 11% water, 0.7% sodium chloride; reduction titration of azo groups indicated a purity of 83%, and the authors stated that the titration estimate of purity was “probably enhanced by the presence of reducible low molecular weight organic impurities containing the azo group as well as inorganic copper salts”; 3,3′-dimethoxybenzidine was < 1 ppm, and benzidine was ≤ 1 ppm) at a concentration of 0, 1000, 3000, or 10 000 ppm (approximately equal to average daily doses of 0, 120, 360, and 1520 mg/kg body weight (bw) per day for males and 0, 140, 470, and 2050 mg/kg bw per day for females, respectively), for the control group and groups at the lowest, intermediate, and highest dose, respectively, for 104 weeks. In addition, 9 males for the control group and 10 males and 10 females for all the other groups were given feed containing CI Direct Blue 218 at a concentration of 0 (controls), 1000, 3000, or 10 000 ppm, respectively, for interim evaluation at 15 months. Survival of exposed male and female mice was similar to that of the controls. At study termination (104 weeks), survival was 44/50, 46/50, 42/50, and 45/50 in males, and 37/49, 40/50, 46/49, and 38/49 in females, for the control group and groups at the lowest, intermediate, and highest dose, respectively. The final mean body weight of mice at the highest dose was 10 000 ppm was 19% lower than that of the control mice for males and 27% lower than that of the control mice for females. Complete necropsies and full histopathological examinations were performed.

In male mice, there was a significant positive trend in the incidence of hepatocellular adenoma, hepatocellular carcinoma, and hepatocellular adenoma or carcinoma (combined) (P < 0.001, logistic regression trend test), with a significant increase in the incidence of hepatocellular adenoma, hepatocellular carcinoma, and hepatocellular adenoma or carcinoma (combined) at the highest dose (P < 0.001; P = 0.019; P < 0.001, respectively, logistic regression test). Renal tubule adenomas were observed in two males at the lowest dose, one male at the intermediate dose, and one male at the highest dose. In addition, one renal tubule carcinoma was observed in another male at the lowest dose. No renal tubule tumours were observed in mice in the control group. The incidence of renal tubule adenoma or carcinoma (combined) in historical controls was 4/1366 (0.3 ± 0.7%; range, 0–2%), and only one renal tubule carcinoma was observed. [The Working Group noted that although there was no statistically significant increase, the renal tubule tumours may have been treatment-related because these tumours are rare in this strain of mouse.] Carcinomas of the small intestine occurred in three males at the highest dose and in one mouse in the control group. In addition, in the 15-month interim evaluation experiment, one carcinoma of the small intestine was observed in a male at the highest dose, but none were reported for the controls. In the historical controls, the incidence of carcinoma of the small intestine was 7/1374 (0.5 ± 1.0%; range, 0–4%).

In female mice, there was a significant positive trend in the incidence of hepatocellular adenoma (P < 0.001, logistic regression trend test), hepatocellular carcinoma (P = 0.012, logistic regression trend test), and hepatocellular adenoma or carcinoma (combined) (P < 0.001, logistic regression trend test), with a significant increase in the incidence of hepatocellular adenoma (intermediate dose, P = 0.041; highest dose, P < 0.001; logistic regression tests) and hepatocellular adenoma or carcinoma (combined) (intermediate dose, P = 0.045; highest dose, P < 0.001; logistic regression tests).

Regarding non-neoplastic lesions, there was a significant increase in the incidence of eosinophilic foci of the liver at the highest dose in males, and a significant increase in the incidence of clear cell foci, eosinophilic foci, and foci (all) of the liver at the highest dose in females.

[The Working Group noted this was a well-conducted study that complied with GLP, males and females were used, and the duration of exposure and observation was adequate. The impurity of the compound used was a weakness of the study. The Working Group noted that, of the impurities, benzidine and 3,3′-dimethoxybenzidine were present at low levels.]

Oral administration (feed)

In a study of chronic toxicity and carcinogenicity that complied with GLP, four groups of 50–51 male and four groups of 50–51 female F344/N rats (age, 6–7 weeks) were given feed containing CI Direct Blue 218 at concentrations of 0, 1000, 3000, or 10 000 ppm (representing average daily doses of approximately 0, 40, 120, and 440 mg/kg bw per day for males, and 0, 50, 140, and 470 mg/kg bw per day for females, respectively) for the control group and groups at the lowest, intermediate, and highest dose, respectively, for 103 weeks (NTP, 1994). The agent was a desalted commercial dye of 60% copper complex of 3,3′-[(3,3′-dihydroxy[1,1′-biphenyl]-4,4′-diyl)bis(azo)]bis[5-amino-4-hydroxy-2,7-naphthalenedisulfonic acid] tetrasodium salt; 11% water, 0.7% sodium chloride; reduction titration of azo groups indicated a purity of 83%, and the authors stated that the titration estimate of purity was probably “enhanced by the presence of reducible low molecular weight organic impurities containing the azo group as well as inorganic copper salts”; 3,3′-dimethoxybenzidine was < 1 ppm, and benzidine was ≤ 1 ppm). Survival of females at the highest dose was slightly reduced, but not significantly lower than that of the control group. At study termination, survival was 30/50, 25/50, 29/50, and 24/51 for males, and 35/51, 29/51, 31/50, and 25/50 for females for the control group and groups at the lowest, intermediate, and highest dose, respectively. Final mean body weights of rats at the highest dose was 11% lower than that of the controls for males, and 9% lower than that of the controls for females. Complete necropsies and full histopathological examinations were performed.

In male rats, there was a significant positive trend in the incidence of squamous cell papilloma of the oral epithelium (pharynx) (P < 0.001, logistic regression trend test), with a significant increase in incidence at the highest dose (P = 0.026, logistic regression test). In addition, one squamous cell carcinoma of the oral epithelium (pharynx) was also observed in one rat at the highest dose; no squamous cell carcinomas of the oral epithelium were observed in 1253 male historical controls. There was a significant positive trend in the incidence of squamous cell papilloma or carcinoma (combined) of the oral epithelium (pharynx) (P < 0.001, logistic regression trend test), with a significant increase in incidence at the highest dose (P = 0.013, logistic regression test). While there was no significant increase in the incidence of forestomach squamous cell papilloma or carcinoma (combined), the incidence at the intermediate dose – control group, 0/50; lowest dose, 0/50; intermediate dose, 3/50 (6%); and highest dose, 1/50 (2%) – exceeded the upper bound of the range observed in historical controls in this laboratory (4/1253, 0.3 ± 0.8%; range, 0–2%). [Since the incidence of basal cell hyperplasia in the forestomach was significantly increased at the two higher doses in males, and forestomach tumours are rare in this strain of rat, the Working Group considered that the increased incidence of forestomach tumours may have been treatment-related.]

In female rats, the incidence of uterine endometrial stromal polyps – control group, 1/50 (2%); lowest dose, 12/50 (24%); intermediate dose, 10/50 (20%); highest dose, 10/50 (20%) – was significantly increased in each treated group compared with controls (lowest dose, P < 0.001; intermediate dose, P = 0.004; highest dose, P = 0.001, logistic regression test). However, there was no dose–response relationship and, when compared with the incidence and range in historical controls (205/1251, 16.4 ± 6.6%; range, 2–30%), the incidence in the concurrent controls (1/50, 2%) was very low. [The Working Group considered that the higher incidence of endometrial stromal polyps in the exposed groups was probably not treatment-related.]

Regarding non-neoplastic lesions, there was a significant increase in the incidence of basal cell hyperplasia of the forestomach in males at the intermediate and highest dose.

[The Working Group noted this was a well-conducted study that complied with GLP, males and females were used, and the duration of exposure and observation was adequate. The impurity of the compound used was a weakness of the study. The Working Group noted that, of the impurities, benzidine and 3,3′-dimethoxybenzidine were present at low levels.]

4.2.1. Is genotoxic

(b) Experimental systems

(i) Non-human mammals in vivo

No data were available to the Working Group.

(ii) Non-human mammals in vitro

See Table 4.1.

Table 4.1

Genetic and related effects of CI Direct Blue 218 in non-human mammals in vitro.

CI Direct Blue 218 induced mutations in the presence, but not in the absence, of metabolic activation in a mouse lymphoma assay (Mitchell et al., 1997). Negative results were obtained when the ability of CI Direct Blue 218 to induce chromosomal aberrations was tested in Chinese hamster ovary cells with and without metabolic activation (NTP, 2020). It significantly increased the frequency of sister-chromatid exchanges compared with the negative control without, but not with, metabolic activation. [The Working Group noted that no information on purity was provided in these studies.]

(iii) Non-mammalian experimental systems

See Table 4.2.

Table 4.2

Genetic and related effects of CI Direct Blue 218 in non-mammalian experimental systems.

Woodruff et al. (1985) reported that CI Direct Blue 218 did not induce sex-linked recessive mutation in meiotic and postmeiotic germ cell stages of Canton-S male Drosophila melanogaster treated with CI Direct Blue 218 in the feed or via injection. [The Working Group noted that no information on purity was provided.]

Results were largely negative in the available Salmonella/microsome mutagenicity assays. Gregory et al. (1981) reported negative results in tests performed under different reductive conditions (addition of riboflavin, nitrogen gas, and sodium dithionate) in the presence of metabolic activation in strains TA98 and TA100. In tests performed in the presence of sodium dithionate in strain TA100, more than double the numbers of revertants per plate in comparison with the negative control were obtained at the three lowest doses tested, whereas results at the two highest doses were comparable with those for the negative controls. [The Working Group noted that the experiment was not repeated, so no conclusion on the mutagenicity of CI Direct Blue 218 in TA100 could be drawn. No information on purity was provided.]

Prival et al. (1984) reported negative results for CI Direct Blue 218 (commercial samples) [the Working Group noted that no information on purity was provided] in the TA98 strain in the presence of metabolic activation (hamster liver S9), with and without modifications to promote the reduction of the azo bond (Prival & Mitchell, 1982). These modifications consisted of a 30-minute preincubation without agitation (to reduce oxygen) and the addition of flavin mononucleotide. ortho-Dianisidine (3,3′-dimethoxybenzidine), the parental aromatic amine of CI Direct Blue 218, was tested for mutagenicity under the same conditions and provided a clear positive response that was unchanged by the addition of CI Direct Blue 218. [The Working Group noted that this could indicate that the dye did not release ortho-dianisidine (3,3′-dimethoxybenzidine) after incubation with S9, because if this was the case, the test with the dye would provide a mutagenic response. In fact, the cleavage of the azo bond of CI Direct Blue 218 would generate 3,3′-dihydroxybenzidine, which gives negative results when tested with S9 at 10%. 3,3′-Dihydroxybenzidine is only mutagenic when S9 is present at 30% (NTP, 2018).] Prival et al. (1984) also reported negative results for CI Direct Blue 218 in strains TA100 and TA1537, and confirmed the negative results for TA98 using rat liver S9.

Reid et al. (1984) reported negative results for a commercial sample of CI Direct Blue 218 in strain TA1538 with endogenous metabolic activation. They tested the dye with and without incubation with a washed suspension of rat caecal bacteria (caecal reduction system) to promote reduction of the azo bond, simulating intestinal metabolism. The assay was also performed with flavin mononucleotide and Syrian golden hamster S9 using the protocol described by Prival & Mitchell (1982). [The Working Group noted that no information on purity was provided.]

Mortelmans et al. (1986) reported negative results for CI Direct Blue 218 (purity, 44.8%) in strains TA1535, TA1537, TA97, TA98, and TA100, with and without endogenous metabolic activation and preincubation. [The Working Group noted that the test was performed under oxidative conditions.]

[The Working Group noted that although the available Salmonella tests were performed with the five strains recommended by the Organisation for Economic Co-operation and Development (OECD, 2020), CI Direct Blue 218 was not tested in one of the strains sensitive to aromatic amines, e.g. YG1041 (Hagiwara et al., 1993). Zwarg et al. (2018) and Umbuzeiro et al. (2021) have shown the importance of including this strain to detect the mutagenicity of azo dyes containing NH₂ radicals, which is the case with CI Direct Blue 218.]

4.2.2. Causes immortalization

Matthews et al. (1993) reported equivocal results for commercial CI Direct Blue 218 in a study with the objective of introducing an improved method for detecting chemically induced morphological transformation of A31-1-13 BALB/c-3T3 cells. CI Direct Blue 218 was tested at six concentrations (125−2000 µM). No concentration–response relationship was observed, but at 250 and 500 µM, significant differences were observed in comparison with the negative control. [The Working Group noted that no information on purity was provided.]