Frits, chemicals

Endpoint summary

Key value for chemical safety assessment

Effects on fertility

Description of key information

A summary document is attached in sections 7.8.1, 7.8.2, and 7.10.2 of IUCLID and in Appendix B6 of the CSR discussing human assessments of reproductive impairment associated with soluble nickel exposures as a worst case scenario. In summary, a reproductive study of female refinery workers in Russia has not demonstrated an association between relatively high soluble nickel exposures (worst case scenario with higher blood and urinary levels) and the following reproductive outcomes: genital malformations (hypospadias and cryptorchidism), spontaneous abortions, small-for-gestational-age newborns, and skeletal malformations (Vaktskjold et al., 2006, 2007, 2008a&b).

Additional information

Data requirements for nickel oxide are fulfilled by reading across the negative results from fertility studies with soluble nickel compounds.

There is no higher-tier reliable study examining fertility endpoints with nickel oxide. However, fertility impairment due to oral or inhalation exposure to nickel compounds (including the most bioavailable forms) has been extensively studied in reliable studies, and no effects on fertility have been found.  Since the reproductive toxicity effects are related to the nickel ion, the lack of fertility effects following exposure to water-soluble nickel compounds (which have higher bioavailability) is relevant to water-insoluble nickel compounds (which have lower bioavailability). If adverse fertility effects are not observed with the most-bioavailable form, they will not be observed with less-bioavailable forms. There are several reliable 13 week and one- and two-generation studies utilizing inhalation or oral administration of water soluble nickel compounds in rats that have not indicated adverse effects on fertility, estrous cycling, sperm parameters, vaginal cytology, copulation and fertility indices, precoital intervals, gestation lengths, gross necropsy findings and histopathology (Ambrose et al., 1976; Smith et al., 1993; RTI, 1988a,b; NTP, 1996; SLI, 2000a,b). While water-soluble nickel compounds are classified as Repro Cat 1B due to developmental effects in rats, neither nickel metal nor any of the insoluble nickel compounds carry harmonized classifications for fertility effects (see Part 3 of Annex VI; CLP Regulation, 2009).

The following information is taken into account for any hazard / risk assessment:

Lack of fertility effects observed in reproductive toxicity studies with soluble Ni compounds (Ambrose et al., 1976; Smith et al., 1993; RTI, 1988a,b; NTP, 1996; SLI, 2000a,b) combined with toxicokinetic data (Ishimatsu et al., 1995) provide sufficient justification that nickel oxide should not be considered a reproductive toxicant with regards to fertility effects.

Effects on developmental toxicity

Description of key information

Data from a reproductive toxicity study with Ni sulfate hexahydrate (SLI, 2000b) combined with toxicokinetic data (Ishimatsu et al. 1995) provide sufficient justification that nickel oxide should not be considered a developmental toxicant.

Additional information

Data from a reproductive toxicity studies with Ni sulphate (SLI, 2000b) and Ni chloride (RTI, 1988a,b) combined with acute oral toxicity (EPSL, 2009a) and toxicodynamic assessments of nickel oxide provide sufficient information regarding reproductive and developmental toxicity for classification. A key factor in the hazard assessment of nickel metal and nickel compounds for reproductive toxicity is the bioavailability of the nickel ion released from the substances following exposure.  A comprehensive read-across program based on bioaccessibility data in synthetic fluids and in vivo acute oral toxicity data has been conducted on a series of Ni substances including Ni oxide (Appendices B1 and B2). While perinatal effects were observed after exposure to soluble Ni compounds, the available toxicokinetic data showed that oral absorption of nickel from insoluble nickel compounds and metallic nickel in rats is many-fold lower than that of soluble nickel compounds (Ishimatsu et al., 1997).Animal toxicokinetic data on nickel oxide and results of a recently completed acute oral toxicity study on nickel oxide suggest that the oral absorption of nickel from nickel oxide is significantly lower than from more soluble forms of nickel and strongly infers that oral exposure to nickel oxide will not allow enough absorption to exceed the threshold for developmental toxicity.

The results (lack of malformations) in a PNDT rat study with a soluble nickel compound (i.e., highest bioavailability) (RTI, 1988a,b) was read-across to nickel oxide to fulfil the reproductive toxicity data requirements for teratogenicity effects under REACH.  In this study, no evidence of teratogenic effects was reported in a multi-generational study where the F2b generation was subjected to a classical teratology assessment (RTI, 1988a,b). The F2b generation was delivered by Caesarean section on gestation day 20 and were examined for external, internal, and skeletal malformations mimicking the protocol of a stand-alone PNDT study, but with extended exposure of the parental animals.  This F2b cohort showed no exposure-related adverse effects. The lack of adverse effects in the F2b generation, examined on gestation day 20, indicated that the developmental effects in the other generational cohorts (i.e., perinatal mortality) were expressed during the perinatal or postnatal periods and not during gestation (RTI, 1988a,b). Rats are considered a sensitive species for studying the reproductive effects of nickel ion, increasing the confidence in the lack of developmental effects observed in the study.

Finally, rat micronucleus studies in bone marrow after repeated oral exposure to a soluble nickel compound did not show adverse effects, even though nickel levels in plasma and bone marrow were increased >30 times and 4 times, respectively, over controls (Oller and Erexson, 2007).  This study confirmed that rapidly dividing tissues (such as those in the developing organism) are not preferentially affected by nickel exposure.

In addition, a single study was identified characterizing developmental toxicity associated with exposure to nickel oxide in rats. No studies characterizing reproductive or teratogenic effects were found in the literature. Weischer et al. (1980) evaluated the effects of a continuous, 21-day gestational inhalation exposure to NiO (unspecified calcining temperature or color) aerosols (0.8, 1.6 and 3.2 mg/m3) in rats. On gestation day 21, fetuses and fetal blood, as well as maternal blood, serum and urine were collected and evaluated for changes in body weight, organ weight, hematological assessments, urinalysis, various clinical chemistry assays, and fetal survival. Exposure related effects including significant reductions in body weight, wet weights of kidneys and lungs, hematocrit, MCV, leukocytes, and erythrocyte count were noted in maternal rats. In fetuses, body weights were reduced and leukocytes and urea in serum were increased. These data suggested that both maternal and developmental toxicities occurred in a dose-dependent fashion following a 21-day gestational inhalation exposure to NiO aerosols. However, because this study only evaluated hematological and clinical chemistry endpoints, only evaluated a single (though relevant) route of exposure, and only evaluated a single species, it is not sufficient to fully characterize potential developmental, reproductive, and/or teratogenic effects associated with exposure to nickel oxide.  Therefore, data from studies conducted with nickel sulfate and nickel chloride will be used for the hazard assessment as described above (RTI, 1988a,b; SLI, 2000b).

A requirement for a second PNDT is waived for all nickel compounds and nickel metal based on the lack of a correlation between nickel exposure and observed (sex organ and skeletal) malformations in the human reproductive studies of nickel-exposed workers reviewed above (Vaktskjold et al., 2006, 2007; 2008a,b). Specific endpoints, their assessment methods, and the number of subjects included: genital malformations in newborns of female nickel-refinery workers were examined with a register-based, nested case-control study (n= 103 cases; 23,038 controls); small-for-gestational-age newborns of female refinery workers exposed to nickel were examined with a register-based, nested case-control study (n= 2,096 cases; 20,740 controls); spontaneous abortions among nickel-exposed female refinery workers were examined with a case-control study (n=184 cases, 1,691 controls); and maternal nickel exposure and congenital musculoskeletal defects were examined with register-based, nested case-control study (n=341 cases, 22,624 controls).  In all studies, nickel exposure was not associated with adverse pregnancy outcome for any of the endpoints examined.  Such large studies provide adequate statistical power to support a conclusion to exclude the risk of reproductive effects (e.g. malformations) from exposure to the chemical (EMA/CHMP Guideline, 2006).  

The following information is taken into account for any hazard / risk assessment:

Data from reproductive toxicity studies with Ni sulphate (SLI, 2000a,b) and Ni chloride hexahydrate (RTI, 1988a,b) combined with toxicokinetic data (Ishimatsu et al. 1995) provide sufficient justification that nickel oxide should not be considered a developmental toxicant.

Toxicity to reproduction: other studies

Additional information

A reproductive study of female refinery workers in Russia has not demonstrated an association between relatively high soluble nickel exposures (worst case scenario for exposure to all nickel substances including nickel metal as it results in higher blood and urinary levels) and the following reproductive outcomes: genital malformations (hypospadias and cryptorchidism), spontaneous abortions, small-for-gestational-age newborns, and skeletal malformations (Vaktskjold et al., 2006, 2007, 2008a&b). Researchers created a birth registry for all births occurring in the region of a Russian nickel refinery at Monchegorsk during the period of the study, which included information on 22,836 newborns and 2,793 pregnancy outcomes surveyed for spontaneous abortions (Vaktskjold et al. 2007; 2008a). They also reconstructed the exposures (using air and urinary nickel measurements) for the female workers at the refineries so as to be able to link specific pregnancy outcomes with occupational exposures via inhalation and systemically bioavailable doses of nickel (i.e. urinary nickel levels).  The study culminated in a series of manuscripts by Vaktskjold et al. (2006, 2007, 2008a,b) describing the results of the investigation. The study demonstrated nickel compound and metal exposure was not associated with adverse pregnancy outcome for 1) male newborns with genital malformations, 2) spontaneous abortions, 3) small-for-gestational-age newborns, or 4) musculoskeletal effects in newborns of female refinery workers exposed to nickel. The data in these manuscripts showed no correlation between nickel exposures (urinary levels as high as 30-fold over background ) and observed reproductive impairment.  It is important to note that genital malformations are considered as one of the most sensitive endpoints for human developmental toxicity while spontaneous abortion in humans would most closely approximate the observation of perinatal lethality associated with nickel exposure in rodents. Further evidence that nickel compound and metal exposure was not adversely affecting the reproduction of these women was provided by the lack of a “small-for-gestational-age” finding and also the lack of an association of male genital malformations with nickel exposure. Both of these findings are considered “sentinel” effects (i.e., sensitive endpoints) for reproductive toxicity in humans.

The work by Vaktskjold et al. (2006, 2007, 2008a,b) is important in demonstrating that no hazard for reproductive impairment from nickel compound and metal exposure exists under the conditions of the study, which were extremely high exposure levels which are no longer present in current refinery operations.  Thus, the reproductive effects observed in rats 1) may not be relevant to humans or 2) may not have been observed in the exposed human population because even the highest achievable female workers’ exposure (179 µg Ni/l in urine) is lower than those achieved in rats at the LOAEL for reproductive effects (2300 µg Ni/L in urine).  In either case, for classification and risk assessment purposes, the relevance of the positive results in rats with soluble nickel compounds (at what seem to be unachievable human exposures) needs to be considered together with the negative results in human studies at Monchegorsk (for the highest exposed human population) in a weight of evidence approach.  At the very least, the data indicates that humans do not appear to be greatly more sensitive to reproductive effects of nickel ion than rats. While the exact mode of action for the perinatal mortality effects of Ni ion are not known, the existing data demonstrates that these are clearly threshold-mediated effects. Because the study correlated the systemically available nickel levels to reproductive outcomes, the lack of reproductive toxicity effects is relevant to nickel compounds and nickel metal. Importantly, the results from human studies showing no reproductive toxicity effects due to nickel compound and metal exposure reported by Vaktskjold et al. were not available at the time when nickel compounds were classified as Cat. 1B reproductive toxicants.

The mode of action for the reproductive toxicity of soluble nickel ion observed in rodents is not currently known; thus its relevancy for humans is also unknown. What is known is that epidemiological studies of female workers exposed to the highest attainable levels of water soluble nickel compounds (via inhalation) and who had urinary nickel levels up to 30-fold above background failed to show an association between exposures to nickel and observed adverse reproductive effects. To place the animal and human results in context, we can compare the urine nickel levels in the workers’ cohort with the urine levels in rat reproductive studies. Background urinary nickel levels in the female Monchegorsk population had a geometric mean of 5.9 µg/l, the low exposure refinery workers had urinary levels up to 70 µg/l (~12-fold increase in urinary levels) and the high exposure workers had urinary levels between 70 and 179 µg/l (up to 30-fold increase in urinary levels). Urinary levels of 70 µg/l in low exposure workers corresponded to approximately 160 µg soluble nickel inhalable exposure/m3 (>1000-fold over ambient air levels).

Comparison of the exposures in female workers to the exposure of rats can be made.  In a rat oral study with nickel sulfate (100% bioaccessible nickel), blood and urinary nickel levels were measured after two years of exposure to 2.2 to 10 mg Ni/kg (Heim et al., 2007; Rush, 2005). A linear dose-response between oral intake of nickel and urinary nickel levels was found. An exposure of 2.2 mg Ni/kg corresponded to a mean urine value of 2300 µg Ni/L (males + females) with blood peak levels of ~70 µg Ni/l. Thus, the rat urinary nickel level at the LOAEL for reproductive developmental effects (2.2 mg Ni/kg) is 33-fold and 13-fold higher than those measured in Low and High exposure nickel refinery workers, respectively. Likewise, the rat urinary nickel level at the NOAEL for developmental effects (1.1 mg Ni/kg) are expected to be 16.5-fold and 6.5-fold higher than those measured in Low and High exposure nickel refinery workers, respectively.  

Thus, the reproductive effects observed in rats 1) may not be relevant to humans or 2) may not have been observed in the exposed human population because even the highest achievable human exposures are lower than those at which adverse reproductive effects have been observed in rats.  In either case, the relevance of the positive results in rats (at unachievable human exposures) needs to be considered together with the negative results in human studies at Monchegorsk (for the highest exposed human population).  

These data show that while a reproductive “hazard” from nickel ion systemic exposure can be demonstrated in animals, this hazard has not been demonstrated in humans.

The full document summarizing these data in humans is attached in Sections 7.8 and 7.10.2 of IUCLID and in Appendix B6 to the CSR.

The following information is taken into account for any hazard / risk assessment:

A summary document is attached in sections 7.8.1, 7.8.2, and 7.10.2 of IUCLID and in Appendix B5 of this CSR discussing human assessments of reproductive impairment associated with soluble nickel exposures as a worst case scenario. In summary, a reproductive study of female refinery workers in Russia has not demonstrated an association between relatively high soluble nickel exposures (worst case scenario with higher blood and urinary levels) and the following reproductive outcomes: genital malformations (hypospadias and cryptorchidism), spontaneous abortions, small-for-gestational-age newborns, and skeletal malformations (Vaktskjold et al., 2006, 2007, 2008a&b).

Justification for classification or non-classification

Ni oxide is not classified for reproductive or developmental toxicity, or effects on fertility according the the 1st ATP to the CLP Regulation. The weight of evidence from all data discussed above demonstrates a consistent lack of sufficient in vivo bioavailability by oral, inhalation and dermal routes, and no classification for reproductive toxicity of nickel oxide is warranted.