Registration Dossier

Administrative data

Description of key information

Additional information

Introductory remark – read-across

 

Read-across entails the use of relevant information from analogous substances (the ‘source’ information) to predict properties for the ‘target’ substance(s) under consideration. Substances whose physicochemical or toxicological properties are likely to be similar or follow a regular pattern as a result of structural similarity may be considered as a category of substances. Structural similarity is a pre-requisite for any read-across approach under REACH (ECHA Read-Across Assessment Framework, 2015).

 

In accordance with Annex XI, 1.5 of the REACH regulation and the ECHA Guidance Read-Across Assessment Framework (ECHA, 2017), the similarities may be based on:

 

1) A common functional group (i.e. chemical similarity within the group);

2) Common precursors and/or likelihood of same breakdown products through physical and/or biological processes which result in structurally-similar degradation products (i.e. similarity through (bio) transformation); or

3) A constant pattern in the changing of the potency of the properties across the group (i.e. of physical-chemical and/or biological properties).

 

Due to the absence of substance specific information for the majority of substances within the cobalt category, the approach will read-across data from representative source substances to all other members of the read-across group.

 

Due to the route-specific toxicological properties of the cobalt category substances, several read-across groups are formed as shown in the table below:

 

 

Route

Read-across group

Cobalt category

oral-systemic

bioavailable cobalt substances group

inorganic poorly soluble

poorly soluble in aqueous solutions with organic ligand

inhalation-local

reactive

non-reactive

 

 

Further details on the read-across approach are given in Appendix 1.1 of the CSR for the oral systemic effects and Appendix 1.2 of the CSR for the inhalation local effects.

 

 

A. Executive summary - Respiratory effects of workers exposed to cobalt and cobalt compounds

 

In the epidemiological studies which were included in this evaluation, occupational asthma was the predominant finding.

Roto (1980) and a follow-up analysis by Sauni et al. (2010) published a detailed survey of lung effects in workers of a Finnish cobalt metallurgical plant in Kokkola. In a cross-sectional part of the study, the occurrence of symptoms of chronic bronchitis and an eventual decrease of ventilatory capacity in workers exposed to cobalt-containing aerosols was determined. The incidence density of cobalt asthma correlated with cobalt exposure levels. However, all diagnosed cases of asthma occurred only under workplace conditions with co-exposures to irritant gases such as sulfur dioxide, hydrogen sulfide or ammonia, whereas in the absence of such irritants in the workplace atmosphere, exposure to cobalt alone at levels of 0.12 mg/m³ (median, min to max 0.02-0.3 mg/m³) did not elicit any asthma-like conditions (Sauni et al., 2010).

 

In a similar cross-sectional study by Swennen et al. (1993), cobalt-exposed workers in a Belgian plant complained of dyspnoea and wheezing, and a concentration-effect relationship was found for the reduction of the FEV1/VC ratio, despite that the average lung function tests were not significantly different between cobalt exposed workers and controls. Verougstraete et al. (2004) examined in a follow-up study the influence of cobalt exposure on lung function parameters in a 13-year health surveillance program (1988-2001). A total of 122 male workers in this plant who had at least 4 pulmonary function tests during the follow-up period (a minimum of three between 1988 and 2001, and one in 2001-2002) were assessed longitudinally. The predominant finding of the longitudinal survey was that cobalt exposure contributed to a decline in FEV1 over time, but only in association with smoking. This confirms the observations made in the Roto (1980) and Sauni et al. (2010) publications that individuals with co-exposure to respiratory irritating chemicals show lung effects at lower cobalt exposure concentrations.

 

Based on the findings of the epidemiological studies in workers by Swennen et al. (1993) and Verougstraete et al. (2004), Roto (1980) and Sauni et al. (2010) as discussed below, a cobalt concentration of 0.12 mg Co/m³ will be used as NOAEC for setting a DNEL.

 

The two available epidemiological studies in the cobalt production plants in France do not report an increase in lung cancer risk among cobalt production workers. However the relevance of these studies is limited due to the small numbers of subjects examined.

 

 

B. Discussion, respiratory effects in humans

 

The main objective of this evaluation is to analyse the adverse respiratory effects of prolonged inhalation exposure to cobalt and cobalt compounds in occupational settings. The evaluation is restricted to health effects in workers resulting from exposures to cobalt metal, cobalt oxides and inorganic cobalt salts. Sources describing effects from co-exposure to cobalt containing diamond polishing dust, hard metal, cobalt alloys (vitallium, stellite), cobalt zinc silicate, and cobalt aluminate spinels were not included in the evaluation, since synergistic or accumulating effects cannot be excluded (Nemery, 1992, Demedts, 1984, Gheysens, 1985, Gennart and Lauwerys, 1990).

The references on the induction of lung cancer in workers are assessed in separate sub-chapters.

Further references on other effects on humans following cobalt exposure, such as skin sensitisation or systemic effects following inhalation exposure are summarised under the relevant sub-chapters of this dossier.

 

 B.1 References on respiratory effects of workers exposed to cobalt and cobalt compounds:

 

Roto (1980) published a detailed survey of lung effects in workers of a Finnish cobalt metallurgical plant in Kokkola that started operations in 1966. The study was carried out between February 1 and November 1, 1977. In a case-control study with 21 cases (male workers with bronchial asthma) and 55 controls (sex and age matched), the risk of asthma was increased for subjects exposed to cobalt (age adjusted OR=4.8, 95%CI=2.0-11.7), i. e. for those workers with exposure to cobalt sulfate or cobalt metal dust for at least six months, compared to non-exposed workers. Smoking was not associated with asthma. The exposure levels ranged from less than 10 to 100 µg Co/m³ in the cobalt plant (stationary sampling) and from 10 to 50 µg Co/m³ in the roasting area (personal sampling). Among workers exposed to cobalt metal and/or cobalt salts, 15 cases of asthma were diagnosed. Five out of them had a positive reaction to CoCl2 in a provocation test and one of them had a positive reaction to dust from the cobalt roasting building. In 12 of the asthmatic cobalt workers, asthma disappeared after removal from exposed workplaces.

In addition, in a cross-sectional part of the study, the occurrence of symptoms of chronic bronchitis and an eventual decrease of ventilatory capacity in workers exposed to cobalt-containing aerosols was determined. Asthmatic subjects were excluded from the cross-sectional study. The population of 224 cobalt workers exposed to cobalt metal, oxides and salts at concentrations in the range from 10 to about 50 µg Co/m³ showed significantly more wheezing than in the reference population of 161 non-exposed subjects, as assessed with a questionnaire (Roto 1980). These findings were predominantly observed in the presence of irritant gases. The chronic production of phlegm and wheezing were clearly associated with smoking among the cobalt workers. The prevalence of chronic bronchitis was 2% in the cobalt exposed workers and 0% among the reference group. Multiple regression analyses indicated that exposure to the air of the cobalt plant did not significantly decrease the FEV1 or FVC of workers when the exposure level was below 100 µg Co/m³. The author concluded that workers exposed to air containing cobalt sulfate at concentrations below 100 µg Co/m³ for 6 to 8 years did not show any increased risk of developing chronic bronchitis. However, smoking highly significantly decreased the FEV1 of cobalt workers.

 

In a later study on the occurrence of respiratory symptoms or diseases in the same Finnish plant, the study population encompassed 110 current and former male cobalt workers who had worked more than 10 years in the cobalt plant (Linna et al., 2003). The reference group consisted of 76 plant employees (68 white-collar workers) and 64 male blue-collar maintenance workers who had worked at least 10 years, but not in the cobalt factory, and who had not been exposed to harmful dusts or fumes. Among the exposed workers (mean cumulative exposure to cobalt 1000 µg/year), there was a significantly increased prevalence of suspected work-related asthma (15 subjects), phlegm, cough with wheezing, shortness of breath with wheezing and breathlessness on exertion than among controls. No chronic respiratory diseases, except asthma, were found among non-smoking cobalt production workers. FEV1 and the respiratory flow rates MEF25 and MEF50 were significantly lower among exposed smokers compared to smoking controls. One new case of occupational asthma (cobalt) with positive reaction in a provocation test and one case of allergic asthma were diagnosed. At concentrations lower than 100 µg Co/m³ cobalt metal or cobalt sulfate exposure increased the risk of asthma by about five times in exposed workers. However, one has to notice constrictively that all cases of cobalt asthma diagnosed referred to workplace exposure conditions where additional irritant gases like sulfur dioxide, hydrogen sulfide or ammonia were present in the ambient air in addition to cobalt.

 

In a follow-up study (Sauni et al., 2010) all the cases of occupational asthma encountered in the Kokkola cobalt plant and diagnosed in the Finnish Institute of Occupational Health were re-evaluated. The analysis was based on the clinical data at the time of diagnosis and during a follow-up visit 6 months later. In addition, the incidence of cobalt asthma in different departments was evaluated on the basis of data on occupational exposures. The significance of exposure to cobalt and to irritant gases in the workplace air in relation to the risk of cobalt asthma was also evaluated based on lung function testing. Between 1967 and 2003, a total of 22 cases of cobalt asthma about 700 cobalt-exposed workers were diagnosed in the cobalt plant that started operating in 1966. All patients except one were male. None of them had positive reactions against cobalt in skin prick tests, indicating a non-immunologic mechanism. The mean duration of symptoms was 7.4 years on average before diagnosis of occupational asthma. Mostly late or dual asthmatic reactions were observed in specific bronchial challenge tests with cobalt which may also suggest a non-immunologic mechanism of asthma. The incidence of cobalt asthma correlated with the exposure levels of cobalt in the corresponding departments of the plant during 1967-1987. The incidence density of cobalt asthma (number of new cases per person-year) was highest in the reduction and powder production department (0.02), where the cobalt exposure levels were highest (median 150 µg/m³, min to max 100 -400 µg/m³). The incidence density in the sulfatising roasting department and leaching and solution purification department was 0.006 and 0.005, respectively, with median exposure levels to cobalt of 100 µg/m³ (min to max 6 -1000 µg/m³) and 30 µg/m³ (min to max 10 -100 µg/m³), respectively. There was significant individual variation in the working time before the onset of symptoms (0.1-17 years). The shortest latencies were in the sulfatising roasting department, where the total dust concentration and the sulfur dioxide level were high (mean exposure to dust of 8.5 mg/m³ and to sulfur dioxide of 1.4 ppm, cf. Linna et al., 2003). All cases of cobalt asthma were encountered in those departments where additional irritant gases like sulfur dioxide, hydrogen sulfide or ammonia were present in the ambient air in addition to cobalt. No cases of cobalt asthma were reported in the chemical department with a cobalt exposure level 120 µg/m³ (median, min to max 20-300 µg/m³) where co-exposure to additional irritant gases was not present. At the time of the follow-up examination 6 months later, non-specific bronchial hyper-reactivity had mostly remained at the same level or increased. In conclusion, the current evidence indicates that as the mean exposure levels to cobalt increase, the risk of occupational asthma induced by cobalt also increases and irritant gases contribute to the risk. When exposure to the causative agent ceases, the symptoms and bronchial hyper-reactivity persisted in some cases.

 

Swennen et al. (1993) carried out a cross-sectional study among 82 male workers exposed in a Belgian cobalt refinery to cobalt dusts. The workers had been exposed for 8.0 years on average (range 0.3 – 39.4 years). Results were compared with those in a sex and age matched group of 82 control workers neither exposed to lung irritants nor to cobalt (air concentrations were below 0.05 µg Co/m³). The workers were exposed to cobalt metal, oxides and salts at concentrations between 2-7700 µg Co/m³. The geometric mean time weighted average (TWA) assessed with personal samplers (total dust) was 125 µg/m³, but for about 25% of the workers, the TWA exceeded 500 µg/m³. The concentrations of cobalt in blood and urine after the shift were significantly correlated with those in air. The concentration of cobalt in urine increased during the workweek. Statistically significant decreases of T3 levels and of some haematological parameters (red blood cell count, haemoglobin content) were found in the cobalt exposed group whereas white blood cell counts were significantly higher. The prevalence of abnormal values of T3, T4, TSH, RBC, and WBC were significantly higher in the cobalt exposed workers than in the controls. The exposed workers had significantly more skin lesions (eczema, erythema) than control workers and complained more often of dyspnoea and wheezing, especially the smokers. In the analysis for a dose-response relationship only dyspnoea during exercise was found to be related to current concentrations of cobalt in air in a logistic regression model. However, results were only presented in graph in which no differentiation between smoker and non-smoker was made. In view of the significant differences in the prevalence of dyspnoea between smokers and non-smokers (as reported in table 2 of that reference) this evaluation might be questioned – a differentiation between the different exposure groups would be more appropriate. It is stated that a significant relation was also found in the exposed group between the reduction of the FEV1/VC and the intensity of the current exposure to cobalt (cobalt in air or urine). However, detailed values of the reduced lung function parameters were not given. No difference between lung volumes, ventilatory performances, carbon monoxide diffusing capacity and serum myocardial creatine kinase was found between the cobalt exposed workers and the control group. No lung abnormalities were detected on chest radiographs in both groups. The authors concluded that the probability of respiratory complaints (dyspnoea) seems to be low (10% or less) when the exposures to cobalt alone does not exceed 50 µg Co/m³. The results also suggest that exposure to high airborne concentrations of cobalt did not cause pulmonary fibrosis.

 

Verougstraete et al. (2004) examined in a follow-up study the influence of cobalt exposure on lung function parameters in workers from the Belgian cobalt-producing plant in a 13-year health surveillance program (1988-2001). A total of 122 male workers in this plant who had at least 4 pulmonary function tests during the follow-up period (a minimum of three between 1988 and 2001, and one in 2001-2002) were assessed longitudinally. Cobalt exposure significantly decreased over the follow-up period, as reflected by measurements in air (personal sampling) and urine. Cobalt dust levels decreased in all three exposure patterns (dry-stage, wet-stage, mixed exposure), but the decrease was sharpest in dry-stage exposure, which had the highest cobalt air values peaking at 1000 µg/m³ at the beginning of the 1990s (no further details are given). When considering the whole cohort, the mean FEV1 was non-significantly lower in smokers than in non-smokers. No significant differences were found in lung-function values of the three patterns of exposure (dry-area, wet-area, mixed exposure). The main finding of the longitudinal survey was that cobalt exposure, as documented by urinary cobalt levels, contributed to a decline in FEV1 over time, but only in association with smoking. No influence of cobalturia on FVC was observed. A weakening of the association between exposure and health effects could have occurred because plant health personnel removed workers from exposure who had high urinary cobalt levels or abnormal pulmonary function studies. The findings of the follow-up are consistent with previous observations by Swennen et al. (1993) that detected a slight deterioration of the FCV1 to VC ratio associated with exposure to cobalt.

 

Morgan (1983) conducted a cross-sectional study among 49 male workers who were exposed in a cobalt production facility. Personal sampling for cobalt ranged from 100-3000 µg Co/m³ with a mean of 520 µg Co/m³ as cobalt oxides and salts. The mean exposure duration was 10.7 years. Lung functions (FEV1, FCV) were not impaired in the cobalt-exposed workers as compared to the referent group of 46 nickel refinery workers who were not exposed to cobalt and were matched for smoking. There were no X-ray changes suggestive of hard metal disease or pneumoconiosis. Furthermore, there were no ECG changes suggestive of cardiomyopathy.

 

 

Summary B1: respiratory effects of workers exposed to cobalt and cobalt compounds:

 

Based on the above described studies in workers exposed to cobalt metal, oxides and salts in the Finnish cobalt plant (Roto et al., 1980 and Sauni et al., 2010), exposures to cobalt at levels of 6-1,000 µg Co/m³ was reported to induce occupational asthma. The symptoms reported after occupational inhalation of inorganic cobalt compounds included diminished pulmonary function, increased frequencies of phlegm, cough, wheezing, and dyspnoea. However, all cases of cobalt asthma that were diagnosed, were in fact related to workplace exposure conditions where irritant gases like sulfur dioxide, hydrogen sulfide or ammonia were also present in the ambient air or the workers had smoking habits in addition to cobalt. When irritant gases were not present in the occupational atmosphere no cases of cobalt asthma were observed at a median cobalt exposure level 120 µg/m³ (min to max 20 -300 µg/m³). It should also be noted that the authors reported that none of the workers with occupational asthma showed a positive skin prick-test result and most showed late or dual reactions in a bronchial challenge test.

 

In addition, occupational exposure to cobalt compounds (2- 7700 µg/m³ geometric mean TWA 125 µg/m³ and in 25% > 500 µg/m³) in a Belgian cobalt refinery was reported to be associated with a higher incidence of dyspnoea and wheezing as well allergic contact-dermatitis, manifesting as eczema and erythema compared to a non-exposed control population (Swennen et al., 1993 and Verougstraete et al., 2004). In this study no difference in respiratory volumes, ventilatory performance, carbon monoxide diffusion capacity or lung abnormalities in chest radiography were observed between the cobalt exposed and the control population. The reduction decline in FEV1 in the cobalt exposed worker population, as documented by urinary cobalt levels, was only associated with smoking habits.

 

 

B.2 Carcinogenic effects of workers exposed to cobalt and cobalt compounds

 

Two studies of one cohort of occupationally exposed workers in a French electrochemical plant, have evaluated the effects of cobalt, cobalt oxides and cobalt salts on possible carcinogenicity in humans:

Mur et al. (1987) studied the mortality in a cohort of 1,143 workers at an electrochemical plant in France which produced cobalt, cobalt oxides, cobalt salts and sodium. The cohort included all men who had worked 1 year or more between 1950 and 1980. According to the employment in different work areas the following subgroups were defined: cobalt production, sodium production, other chemicals production, maintenance services, and general services. Those groups were mutually exclusive, i.e., only the subjects belonging to one category exclusively were taken into account. Exposure levels of cobalt were not reported. The mortality data of the French national population were used as reference. Among cobalt production workers, there was a relative increase in deaths from cancers of the trachea, lung and bronchus (SMR 4.66; 95%CI 1.46-10.64 based on four cases). No corrections for cigarette smoking were possible. The relationship between cobalt production and lung cancer mortality seemed to be supported by a case-control analysis nested in the cohort study. Among cases (deaths from lung cancer) there were 44% of workers who had only ever been ever employed at the cobalt production (all for more than 10 years), there were only 17% among the controls. However, the difference was not statistically significant. The authenticity of the occupational origin of this risk could not be established due to the low number of cases and because the role of smoking and of simultaneous exposure to arsenic and nickel could not be taken into account.

 

Moulin et al. (1993) made a follow-up study of the same population, extending the observation period from 1980 to 1988. The total cohort comprised 1,148 subjects, i.e., the 1,143 subjects of the former study plus 5 persons who had previously been omitted due to incomplete data. The SMR for all causes of death was 0.85 (95%CI 0.76-0.95) for the whole cohort, and 0.95 (95%CI 0.78-1.26) for the sub-cohort of workers born in France. With regard to lung cancer mortality among cobalt production workers, the SMRs were 0.85 (95%CI 0.18-2.50, 3 cases) for the whole cohort and 1.16 (95%CI 0.24-3.40, 3 cases) for the sub-cohort. Any excess of mortality from diseases of the circulatory and of the respiratory systems did not appear among cobalt production workers. Maintenance workers, however, exhibited a non-significantly elevated SMR for lung cancer (1.80, 95%CI 0.78-3.55), reaching statistical significance for duration of exposure and time since first exposure ≥30 years. This finding could not be clearly explained apart from the fact that asbestos exposure may have occurred. This follow-up study does not support the hypothesis of a relationship between lung cancer and exposure to cobalt, but the significance of the study is limited by the very small number of cases if one considers only subjects having been exposed in the cobalt production.

 

Sauni et al. (2017) performed a retrospective cohort study, made up of all males employed for at least one year at the Kokkola cobalt plant (Freeport Cobalt Oy) during the period 1968–2004. The cohort of 1004 men was identified from the company’s employment records. The study cohort (995 workers) consisted of males only, which were employed by the Finnish cobalt plant for at least a year during 1968–2004 and the cohort was divided into subcohorts by exposure levels (variable exposure with peak exposure, low exposure, moderate exposure, high exposure). Standardised incidence ratios (SIRs) and 95% confidence intervals (95% CIs) were calculated as ratios of the observed numbers of cancer cases and the numbers expected on the basis of incidence rates in the population of the same region. The results suggest that occupational exposure to cobalt is not associated with an increased overall cancer risk (92 cases were diagnosed; SIR = 1.00; CI = 0.81 -1.22) or lung cancer risk (6 cases were diagnosed; SIR = 0.5; CI = 0.18 – 1.08) among cobalt workers. Overall, there was no dose-response effect across the different exposure levels and the incidence of any other cancer type.

 

An international epidemiological study (Kennedy et al., 2017; Westberg et al., 2017ab; Marsh et al., 2017; McElvenny et al., 2017; Wallner et al., 2017; Morfeld et al., 2017) was conducted retrospectively to evaluate mortality among hardmetal production workers that were exposed to cobalt (also partly to tungsten and nickel) during production in view of cancer, especially to lung cancer. The study included 32,354 workers from three companies and 17 manufacturing sites in five countries (eight US sites, three German sites, three Swedish sites, two UK sites, and one Austrian site), each independently conducted under the direction of country-specific occupational epidemiology experts. Job-exposure matrices (JEMs) were constructed for cobalt, tungsten, and nickel over the time period 1952 to 2014. The JEMs consisted of job class categories, based on job titles and processes performed, and exposure estimates calculated from available company industrial hygiene measurements. The pooled analysis of country-specific cohort data from the international study of hardmetal production workers provided no consistent evidence that work in this industry is associated with an increased risk of lung cancer. There is no evidence that duration, average intensity or cumulative exposure to tungsten, cobalt or nickel, at levels experienced by the workers examined, increases lung cancer mortality risks. There is no evidence that work in the US or EU hardmetal industry increases mortality risks from any other cause of death. The results are consistent with the country-specific study findings. In addition, there was no evidence of an occupationally related risk for the NMRD subcategory, such as hardmetal disease and pneumoconioses. This subcategory SMR-R for long-term workers was 0.99 (95% CI=0.76–1.26). This is consistent with the country-specific analyses, which found small numbers of deaths in this NMRD category and no evidence of an exposure-response relationship.

 

Summary B2: carcinogenic effects of workers exposed to cobalt and cobalt compounds

 

A significant database on epidemiological studies in workers of (i) cobalt producing plants or (ii) the hardmetal industry did not find an increase in lung cancer risk among these workers. None of the studies, involving more than 34,000 workers showed increased risks of lung cancer mortality or lung cancer incidence.

 

 

C. Overall evaluation, respiratory effects in humans

 

Several epidemiological studies have been conducted in the past for the assessment of adverse health effects in particular on respiratory function in workers exposed to different concentrations of cobalt metal, oxides and salts under various occupational conditions.

The reported effects of occupational inhalation exposure to inorganic cobalt compounds included reduced pulmonary function, increased frequencies of phlegm, cough, wheezing, and dyspnoea. In addition, exposure to cobalt compounds in humans resulted in allergic dermatitis, manifesting itself as eczema and erythema. However, interstitial lung disease was not found in workers exposed by inhalation to cobalt metal, oxides and salts. No other clinical findings could be related to exposure to cobalt compounds.

 

Based on the most comprehensive and recent evaluation of cases of occupational asthma in workers exposed to cobalt compounds alone, no adverse effects were observed at cobalt exposures of up to 0.12 mg/m³ (min to max 0.02 -0.3 mg/m³). Effects on respiratory function were only observed at lower concentrations when co-exposure to irritant gases occurred. An exposure level of 0.12 mg/m³ can therefore be regarded as a human NOAEC via inhalation for local effects for repeated exposure.

 

The limited numbers of epidemiological studies in the cobalt production plant in France do not report an increase in lung cancer risk among cobalt production workers. However, the significance of the studies was restricted due to the small number of subjects. According to IARC, there is inadequate evidence in humans for the carcinogenicity of cobalt metal without tungsten carbide (IARC 2006).

 

Acute responses of the lung to chemical injury are associated with irritant and inflammatory reactions that may cause changes in airway reactivity and pulmonary oedema on short term exposures. Chronic inflammatory reactions are likely threshold-based responses associated with lung tissue fibrosis, emphysema, asthma, and finally tumour formation. Persistent inflammatory processes may lead to unrestricted cell growth (lung tumours) by a cascade of mechanisms that directly/indirectly damage DNA or by generation of activated-oxygen species which may also elicit indirect damage to DNA and cellular toxicity (see Casarett and Doull, pp 522-526, 2001 for references and discussion).

 

The hallmark of inhalation effects in short and long-term experimental animal exposures to cobalt (+2) substances is inflammation of the respiratory tract (Bucher et al., 1990, NTP TR 471, 1998). Experimental rodent inhalation assays of cobalt sulfate indicated that inflammation of the respiratory tract is already well developed by 16 days in both rats and mice (NTP Report 5 1990). Although direct genotoxicity prior to inflammatory reactions can’t be ruled out, the consequences of inflammatory responses, possibly with the inclusion of indirect effects on DNA, are more plausible mechanisms for the development of experimental lung tumours reported for rodents exposed to cobalt sulfate.

 

In humans, occupational exposures to cobalt substances in the absence of competing metals have been associated with chronic bronchitis (coughing/wheezing, dyspnoea, and phlegm production for >3 weeks) and asthma (reduction of FEV1 by 15%; Swennen et al 1993, Roto, 1980, Linna et al 2003, Suani et al 2010) but not with lung cancer (Moulin et al., 1993; Sauni et al., 2017; Kennedy et al., 2017; Westberg et al., 2017ab; Marsh et al., 2017; McElvenny et al., 2017; Wallner et al., 2017; Morfeld et al., 2017). The reported lung effects may be viewed as indicators of inflammatory reactions and in some cases immune reactions. The health-based endpoint for chronic bronchitis is somewhat subjective because it is questionnaire-based and up to the discretion of the individual. However, the end point for asthma is based on quantitative clinically-compliant lung function tests. Cobalt exposure levels by inhalation, not associated with changes in lung function, might be considered as a health-based point of departure in a quantitative assessment to calculate exposure levels at which inflammatory processes that may include lung carcinogenicity after chronic exposures, are estimated to be low-risk (e.g. < 10-5).  

 

 

D. Selection of a relevant dose descriptor, respiratory effects in humans

 

Whereas it has previously been commented (Lison et al., 2001) that there is some evidence of soluble cobalt(II) cations exerting genotoxic activity in vitro and in vivo in experimental systems but evidence for carcinogenicity in humans is lacking, this statement requires modification based on a recent comprehensive experimental programme involving in vitro gene mutation as well as in vitro and in vivo clastogenicity testing:

 

-      recent guideline-conform, state-of-the-art in-vitro gene mutation assays conducted under GLP for a whole range of cobalt substances have yielded unequivocally negative results

-      in view of the numerous published in-vitro clastogenicity data with predominantly positive outcome, a similar testing programme has also been launched for in-vivo clastogenicity assays, which to date have also yielded negative results.

 

From the above, it may be concluded that cobalt cations do not interact directly with DNA, but indirect reactions such as clastogenic effects cannot be excluded completely. Therefore, it is currently assumed that a (practical) threshold can be identified based on inflammation (e.g. indices of lung function) mechanisms. In humans, no reports of genotoxicity have been associated with cobalt exposure. An attempt is undertaken to identify a relevant dose descriptor as a point of departure for the derivation of a DNEL for local effects via inhalation based on the available epidemiological data in the cobalt industry.

 

In the above mentioned epidemiological studies, occupational asthma was the predominant finding in a case-control study by Roto (1980) and a follow-up analysis by Sauni et al. (2010). However, all diagnosed cases of asthma occurred only under workplace conditions with co-exposures to irritant gases such as sulfur dioxide, hydrogen sulfide or ammonia, whereas in the absence of such irritants in the workplace atmosphere, exposure to cobalt alone at levels of 0.12 mg/m³ (median, min to max 0.02-0.3 g/m³) did not elicit any asthma-like conditions (Sauni et al., 2010).

 

In a similar, cross-sectional study by Swennen et al. (1993) and a follow-up analysis by Verougstraete et al., 2004, cobalt-exposed workers complained of dyspnoea and wheezing, and a concentration-effect relationship was found for the reduction of the FEV1/VC ratio, despite that the average lung function tests were not significantly different between cobalt exposed workers and controls. The reduction in FEV1 in the cobalt exposed worker population was only associated with smoking habits. This confirms the observations made in the Roto (1980) and Sauni et al. (2010) publications that individuals with co-exposure to lung irritating chemicals show lung effects at lower cobalt exposure concentrations.

 

Based on the findings of the epidemiological studies in workers by Swennen et al. (1993) and Verougstraete et al. (2004), Roto (1980) and Sauni et al. (2010) as discussed above, a cobalt concentration of 0.12 mg Co/m³ will be used as NOAEC for the setting of a DNEL for local effects in humans exposed via inhalation.

 

No clinically significant cardiac dysfunction, no evidence of polycythaemia and only equivocal indications of interferences with thyroid metabolism were observed in workers occupationally exposed to inorganic cobalt compounds. Therefore it can be concluded that systemic effects following inhalation exposure are expected at higher dose levels compared to the dose levels for local effects. No point of departure for systemic effects following inhalation exposure will be established. The hazard assessment for systemic effects will be based on animal data following oral exposure. Please refer to chapter 5.6 in the CSR or section 7.5.1 in the IUCLID for further information.

 

Since the human epidemiological data for carcinogenicity via inhalation are considered inadequate for the human health risk assessment under REACH, the available animal data will be used. Please refer to the chapter 5.8 in the CSR or section 7.7 in the IUCLID for further information.

 

 

E. Summary of specific organ effects in humans

 

Effects on cardiac system

 

Thirteen references on effects of cobalt on the cardiac system in humans were identified. Two references were on patients suffering from anaemia under medical therapy with an iron-cobalt preparation (Bianchi 1989, Little 1958). A further eight references were on patients suffering from heart disease, all of which heavily consumed beer containing cobalt as an additive, although a clear causal relationship between the cobalt intake and the heart disease could not be established (Alexander, 1969, 1972, Bonenfant, 1969, Grinvalsky, 1969, Kesteloot, 1968, Morin, 1969, 1971, Sullivan, 1969). One reference was on two patients employed in the mineral assay industry assumed to be exposed via inhalation to mineral dust containing cobalt (among other agents) suffering from congestive heart failure due to non-inflammatory cardiomyopathy (Jarvis, 1992). Another study with volunteers (male and female) ingesting approx. 1.0 mg cobalt/day (0.080 – 0.19 mg Co/kg/day) of a commercially available cobalt supplement (cobalt (II) chloride) over a period of 90 days, showed no clinically significant changes during an echocardiographic examination (Tvermoes et al., 2014). Lastly, one reference was on one female patient having a hip prothesis assumedly containing cobalt (no full description of prothesis was given). Examinations showed that the patient had elevated serum cobalt as well as chromium levels and showed features consistent with a diagnosis of cobalt cardiomyopathy, though myocardial cobalt levels could not be obtained (Khan et al., 2015). The reference was not considered to be relevant for human health risk assessment due to deficiencies in reporting.

 

 

Effects on thyroid

 

Nine references on effects of cobalt on the thyroid in humans were identified. Humans exposed in occupational settings to cobalt containing substances were exposed via inhalation, whereas case reports with pharmaceutical application were via oral exposure. Slight interferences with thyroid metabolism are reported in workers of a cobalt refinery or in painters exposed to an average of 50 µg Co/m³ (as semi-soluble cobalt-zinc-silicate) over about 15 years, but no clinical cases of hypothyroidism were diagnosed. No evidence of abnormalities of the thyroid gland was seen in plate painters exposed to 800 µg Co/m³ for 11 years (Raffn, 1988, Prescott, 1992, Christensen, 1994, Roche, 1956, Chamberlain, 1961). Three case reports were on children (male and female) showing thyroid enlargement which was presumably induced by an oral therapy with a cobalt containing preparation (Washburn, 1964, Klinck, 1955, Kriss, 1955). Another study with volunteers (male and female) receiving approx. 1000 µg cobalt/day (10 – 19 µg Co/kg/day) as cobalt dichloride for a period of 31 days, showed no findings associated with thyroid dysfunction (Finley et al., 2013). Several studies on human exposure were aimed to reveal adverse health effects in workers or in the general population exposed to different concentrations of cobalt under diverse occupational conditions or in patients being treated with cobalt-containing preparations. Only equivocal indications of interferences with thyroid metabolism were observed in workers occupationally exposed to Cobalt.

 

Effects on haematopoietic system

 

Eight references on effects of cobalt on the haematopoietic system in humans were identified. Three case reports are on patients suffering from anaemia being treated orally with cobalt chloride (Gross, 1955, Duckham, 1967, Sederholm, 1968, Taylor, 1977). Pregnant women given cobalt chloride during the third trimester at 0.45 to 0.62 mg Co/kg bw per day did not have increased haemoglobin and red blood cells (Holly 1955). One case report is on the treatment of 6 assumed healthy men (20-47 years of age) with cobalt chloride, subsequently developing polycythaemia (Davis, 1958). A study with volunteers (male and female) receiving approx. 1000 µg cobalt/day (10 – 19 µg Co/kg/day) as cobalt dichloride for a period of 31 days, showed no findings associated with polycythemia (Finley, 2013). Lastly, another study with volunteers (male and female) ingesting approx. 1.0 mg cobalt/day (0.080 – 0.19 mg Co/kg/day) of a commercially available cobalt supplement (cobalt (II) chloride) over a period of 90 days, showed no adverse effects in haematological parameters (Tvermoes, 2014).

 

Neurotoxicity

 

Two references on neurotoxic effects of cobalt in humans were identified. The reference on effects on the neurological system in humans described the effects seen in one male (48 years old), who apparently inhaled raw cobalt powder for 20 months, working 50 hours a week. The symptoms shown by the male were progressive bilateral deafness with tinnitus and visual failure. Blood cobalt level was 234 µg/L 3 months after cessation of work, declining to 14.7 µg/L another 3 months later (Meecham, 1991). Another study with volunteers (male and female) ingesting approx. 1.0 mg cobalt/day (0.080 – 0.19 mg Co/kg/day) of a commercially available cobalt supplement (cobalt (II) chloride) over a period of 90 days, showed no clinically significant changes during audiological, neurological and ophthalmological examinations (Tvermoes, 2014).

 

 

 

Conclusion

 

Several human case reports exist on adverse effects in specific organs in humans potentially associated with cobalt exposure. However, in many cases the persons were exposed to other substances as well or single cases of overexposure with no further information on other confounding factors were reported. Consequently no reliable causal relationship to cobalt exposure can be established. Furthermore a reliable quantification of doses/exposures is not provided in any of these case reports, which is why this information is insufficient for human health hazard assessment. The quality of the information therefore does not comply with the criteria for DNEL/DMEL derivation, as laid down in the ECHA guidance R8 (Appendix R8-15, p148 in conjunction with regulation (EC) 1907/2006, Annex XI).

 

The summary endpoint study records can be found under chapter 5.6.2. Repeated dose toxicity – human information in the CSR or under section 7.10.2. in the IUCLID.