Note: Following the terms of reference, this opinion focuses on specific questions related to human health issues and does not cover ecosystem protection.
Non-Cancer
1. In the light of the published literature and the arguments discussed in the Position Paper what interpretation should be placed on the 0.03 mg Ni/m 3 dose level in the NTP study in rats (using nickel sulphate hexahydrate)?
The CSTEE does not find that the inhalation concentration of 0.03 mg Ni/m 3 represents a NOAEL in rats, since indications of chronic active inflammation and increased lung weights were seen in animals undergoing interim sacrifice.
2. What is the Committee's opinion on the relative sensitivities of rats and humans to exposure by inhalation to soluble nickel compounds? In the Committee´s opinion is a factor of 10 appropriate when extrapolating from rats to humans?
The CSTEE cannot see that there are convincing data supporting the notion that rodents should be more sensitive than humans towards the respiratory effects of soluble nickel compounds. The CSTEE finds it acceptable to use a factor of 10 to extrapolate from experimental animals to humans. However, in the development of a limit value for nickel in ambient air, it should be recognised that soluble nickel species, which are the key contributors to the non-cancer respiratory effects of nickel compounds, do not appear to constitute more than maximally 50 per cent of the total nickel compounds in air as judged from the limited data available. Thus, the CSTEE finds that a limit value of 20 ng Ni/m 3, rather that the value of 10 ng Ni/m 3, is supported by the totality of the existing data.
Cancer
1. In the Committee´s opinion, what chemical forms of nickel in ambient air are the most relevant for the assessment of humans risk and the derivation of a limit value for total nickel
Nickel carcinogenicity will be dependent on the time-integrated intracellular concentration of nickel ions as the active entity, so that the relative potency of various nickel species will be related to their bioavailability and lung burden. In the rat experiments it is the insoluble nickel compounds that are most potent with respect to carcinogenicity. However, epidemiological evidence clearly points to the carcinogenic effect of soluble nickel compounds.
SPECIFIC COMMENTS TO THE POSITION PAPER
Exposure
Species specific measurements of nickel compounds in ambient air (daily means) have been performed at two sites in Dortmund, Germany:
This relative distribution between the various nickel species has been confirmed by measurements in several sites (Feuchtjohann et al., 2000; Broekart, personal communication. See also manuscript by this group submitted to J. Environm. Monit. And PhD-thesis of Feuchtjohann). This indicates that soluble nickel constitutes up to approx. 50% of total nickel in urban air. Therefore, the relative proportion of the various nickel species should be taken into account when establishing limit values for nickel in air ( i.e. a correction for the relative proportion of individual species contributing to an endpoint).
Non-cancer effects
Nickel subsulfide: Chronic active inflammation of the respiratory tract was seen in 2 year inhalation study with rats and mice at all concentrations tested ( > 0.11 mg Ni/m 3).
Nickel oxide: Chronic active inflammation of the respiratory tract was seen in 2 year inhalation study with rats and mice at all concentrations tested ( > 0.5 mg Ni/m 3).
Nickel sulphate hexahydrate: Macrophage hyperplasia was seen in 8 out of 10 female rats and 10 out of 10 male rats at 0.03 mg Ni/m 3 in 13 week inhalation study. An increase in chronic active inflammation at the 7-month interim evaluation of 2 year inhalation study was seen in 4 of 5 male rats (0/5 controls) and 2 of 5 female rats (0/5) controls administered 0.03 mg Ni/m 3. At 15-month interim evaluation there was an increasing trend in the absolute lung weights of both male and female rats. At 2-year termination significant increases in chronic active inflammation and fibrosis were seen in rats and mice at > 0.6 mg Ni/m 3, but not significantly increased at 0.03 mg/m 3. Thus, it can be argued whether 0.03 mg Ni/m 3 should be viewed as a clear NOAEL in rats. In female mice there were clear-cut evidence of chronic active inflammation, bronchialisation and macrophage hyperplasia in the lung at the lowest inhalation concentration of 0.06 mg Ni/m 3 so that this is a LOAEL and a NOAEL was not identified. Such effects were also noted at 0.11 mg Ni/m 3 in male mice.
Genotoxicity
Soluble and insoluble nickel compounds have shown slight, but clearly positive responses in in vitro studies of gene mutations, sister chromatid exchanges, micronuclei (only soluble compounds tested) and cell transformation in mammalian cells (reviewed in ATSDR, 1997). Thus, the CSTEE concludes that a genotoxic mechanism of action for soluble nickel compounds cannot be discounted. There is evidence that the genotoxic effects of nickel compounds may be indirect through inhibition of DNA repair systems (Hartwig, 1998). There is very limited information on the potential for in vivo genotoxicity of nickel compounds.
Carcinogenicity
Nickel oxide, nickel subsulfide and nickel sulphate hexahydrate have been tested by the US National Toxicology Program (NTP) for carcinogenicity in male and female rats and mice for 6 hours/day on 5 days per week for 112 weeks. There was some evidence for carcinogenicity of nickel oxide in male and female rats, clear evidence for carcinogenicity of nickel subsulfide in male and female rats, whereas there were no evidence for carcinogenicity of nickel sulphate hexahydrate in rats. None of the nickel compounds showed carcinogenic potential in mice. The lack of evidence for carcinogenicity of nickel sulphate hexahydrate can be due to the relatively low lung burden that was tested, since the exposure levels had to be kept lower than for nickel oxide and nickel subsulfide due to respiratory toxicity of nickel sulphate hexahydrate. Thus, the lung burden (amount of nickel per g of lung) from the highest exposure concentration of nickel sulphate hexahydrate was approximately 6 times lower than the lowest exposure concentration to nickel subsulfide. In studies with parenteral administration, soluble nickel compounds induce local tumours, albeit with much lower potency than that seen with insoluble nickel compounds. Therefore, the CSTEE concludes that the lack of evidence of carcinogenicity of nickel sulphate hexahydrate in the NTP study cannot be taken as evidence of lack of carcinogenic potential for soluble nickel compounds. Nickel carcinogenicity will be dependent on the time-integrated intracellular concentration of nickel ions as the active entity, so that the relative potency of various nickel species will be related to their bioavailability and lung burden. In the rat experiments it is the insoluble nickel compounds that are most potent with respect to carcinogenicity.
Limit Value Based on Non-Cancer Effects
Since there are differences in opinion about the NOAEL in rats, the Working Group has agreed to use 0.06 mg Ni/m 3 as a LOAEL as the starting point for the risk assessment. The CSTEE considers the increased lung weights noted at 0.03 mg Ni/m 3 in the 15-month interim evaluation indicative of an adverse reaction, so that the 0.03 mg Ni/m 3 does not clearly represent a NOAEL value. The Working Group has used 0.06 mgNi/m 3 as a starting point. In this situation, one would need to apply a larger uncertainty factor than usual for extrapolating from a LOAEL to a NOAEL than is customary ( i.e. a factor of at least 3). Also, this is indicated due to the high percentage of animals that were affected at 0.06 mg Ni/m 3.
An argument has been put forward that humans should be less sensitive than experimental animals towards the respiratory effects of inhaled soluble nickel compounds. This should imply that one should not need a toxicodynamic interspecies uncertainty factor larger than 1. However, the CSTEE cannot see that there are convincing data supporting the notion that rodents should be more sensitive than humans. Thus, a default uncertainty factor for toxicodynamic extrapolation of 3.16 (rather than the conventional 2.5 for organic chemicals) should be applied. Whereas insoluble nickel compounds enter cells via phagocytosis and are retained in the lung tissue for a long time, soluble forms of nickel are inefficiently taken up by cells by the magnesium transport system, but rapidly cleared from the tissues and excreted in the kidneys. However, there are no data to compare the toxicokinetics of soluble nickel compounds between rodents and humans. Thus, also for toxicokinetic extrapolation between experimental animals and humans a default value of 3.16 (rather than the conventional factor of 4 for organic chemicals) can be argued (the position paper uses a default factor of 3).
In the 2-year inhalation studies, the animals were exposed for 6 hours per day and 5 days per week. The conversion factor from this non-continuous exposure to continuous, although only stated as numbers in Table 2.6.6 of the position paper, is (24/6 x 7/5) = 5.6. The Working Group has rounded this off to a conversion factor of 6, which is deemed acceptable.
There does not seem to have been any discussion within the Working Group on applying an interindividual (intraspecies) uncertainty factor of 10. It should be remembered that this factor is meant to cover interindividual variability in both toxicodynamics and toxicokinetics. A major source for interindividual variability in toxicokinetics, is metabolic differences among humans. Such processes would not be relevant for inorganic compounds such as nickel ions. Thus, the default uncertainty factor for interindividual variation in toxicokinetics of 3.16 may be too large for nickel sulphate, although processes other than metabolism could also result in interindividual differences in toxicokinetics. However, since there are no real data on toxicokinetic variability for nickel sulphate in humans, a conservative approach would be to retain the default value. Also, there are no data on interindividual variation in nickel toxicodynamics. Therefore, the CSTEE supports the application of an overall interindividual uncertainty factor of 10.
The Working Group proposes a limit value of 10 ng Ni/m 3 for nickel based on non-cancer effects seen after inhalation exposures to nickel sulphate hexahydrate in rats and mice. This value is reached in the following way: 0.06 mg Ni/m 3 (60 m g Ni/m 3) divided by an uncertainty factor of 10 for LOAEL to NOAEL extrapolation gives 6 m g Ni/m 3. Since this is for non-continuous exposure, dividing by the factor of 6 to arrive at continuous exposure gives a concentration of 1 m g Ni/m 3. When dividing by uncertainty factors of 10 each for interspecies extrapolation and intraspecies variability, the limit value of 10 ng Ni/m 3 is reached.
The CSTEE accepts the composite uncertainty factor of 10 for animal to human extrapolation, which the Working Group has proposed. Although the CSTEE would support the use of the various assessment factors in this calculation, this limit value does not take into account that ambient exposures to nickel are composed of different nickel species with quite large differences in potency with respect to non-cancer effects. From the limited specific exposure measurements, it appears that soluble nickel compounds do not constitute more than maximally 50 per cent of the total nickel compounds in ambient air. Based on this, the CSTEE finds that a limit value of 20 ng Ni/m 3, rather than the value of 10 ng Ni/m 3, is supported by the totality of the existing data.
Limit Value Based on Carcinogenic Effects
In the EU insoluble nickel compounds (nickel oxide, nickel monoxide and nickel sulphide) are classified as Category 1 (known human carcinogens), whereas metallic nickel, slightly soluble and soluble nickel compounds (nickel carbonate, nickel hydroxide, nickel sulphate and nickel tetracarbonyl) are classified as Group 3 (possible carcinogens). IARC has classified nickel compounds (nickel sulphate, and combinations of nickel sulphides and oxides encountered in the nickel refining industry) as carcinogenic to humans (Group 1). The increased risk of lung cancer noted in the Norwegian Falconbridge cohort, implicates an important role of soluble nickel compounds in cancer development (IARC, 1990; Andersen et al., 1996; Grimsrud et al., 2000). The recent updates did not reveal appreciable differences in the risk estimates. Excess risk among Clydach workers seems likely to be due, at least partly, to exposure to soluble nickel (IARC, 1990). Based on an overall evaluation of epidemiological and in vitro and in vivo experimental data, the CSTEE finds that there is sufficient evidence for classifying soluble nickel compounds as known human carcinogens and that a genotoxic component in the mode of carcinogenic action is probable. Thus, the CSTEE does not support the application of a threshold approach for assessing the carcinogenic risks associated with exposure to ambient nickel compounds. As pointed out, increasing evidence has been developed indicating that the genotoxic effects of nickel compounds may be indirect (Hartwig, 1998). This may mean that nickel compounds will show non-linear dose-responses with respect to carcinogenicity, however, based on the available information it is at present not possible to sufficiently evaluate this possibility.
The CSTEE supports the recommendation from WHO (1999) of a unit risk of 3.8 x 10 -4 ( m g Ni/m 3) -1 based on the excess risk seen in the Falconbridge nickel worker cohort. This corresponds to concentrations of 25 ng Ni/m 3 and 2.5 ng Ni/m 3 for increased life-time risks of 1:100,000 and 1:1,000,000, respectively. These estimates are conservative in nature, given the linear extrapolation over many orders of magnitude from the observed excess risk in exposed humans. Also, there are considerable differences in carcinogenic potency among the different nickel species in ambient air, with the most potent sulfidic nickel only constituting up to 10 percent of the sum of nickel species in air as judged from the limited amount of data available. Thus, the CSTEE concludes that the limit value of 20 ng Ni/m 3 proposed for non-cancer effects, also is likely to provide reasonable protection of the general population to the carcinogenic effects of nickel compounds in ambient air.
References
Andersen A, Berge SA, Engeland A & Norseth: Exposure to nickel compounds and smoking relation to incidence of lung and nasal cancer among nickel refinery workers. Occup. Environ. Med. 53, 708-713, 1996.
Feuchtjohann L, Jakubowski N, Gladtke D, Barnowski C, Klockow D & Broekaert JAC: Determination of soluble and insoluble nickel compounds in ambient air dust by graphite furnace atomic absorption spectrometry and inductively coupled plasma mass spectrometry. Fresenius J. Anal. Chem. 366, 142-145, 2000.
Grimsrud TK, Berge SR, Resmann F, Norseth T & Andersen A: Assessment of historical exposures in a nickel refinery in Norway. Scand. J. Work. Environ. Hlth. 26, 338-345, 2000.
Hartwig A: Carcinogenicity of metal compounds: possible role of DNA repair inhibition. Toxicol. Lett. 102-103, 235-239, 1998
WHO. World Health Organisation: Air Quality Guidelines for Europe, 1999.
IARC Monographs, Vol. 49, 1990