94: Cobalt

The name cobalt (Co) originates from kobold, German for “goblin,” and was given to the cobalt containing ore, cobaltite (CoAsS), because it made exposed miners ill. However, the miners’ illness was more likely a result of exposure to arsenic rather than cobalt. Brandt discovered cobalt in 1753 during an attempt to prove that an element other than bismuth gave glass a blue hue.

With an atomic number of 27 and a molecular weight of 58.93 Da, cobalt is a light metal with a melting point of 1768.2°K and a boiling point of 3373°K. These attributes make elemental cobalt (Co⁰) very useful in industry, where it is primarily incorporated into hard, high-speed, high-­temperature cutting tools. When aluminum and nickel are blended with cobalt, an alloy (Alnico) with magnetic properties is formed. Other uses for cobalt include electroplating because of its resistance to oxidation and as an artist’s pigment due to its bright blue color.

A Co³⁺ ion is at the center of cyanocobalamin (vitamin B₁₂), which is synthesized only by microorganisms and is not found in plants. Common dietary sources are fish, eggs, chicken, pork, and seafood. A diet deficient in cyanocobalamin results in pernicious anemia. Hydroxocobalamin, a Co³⁺ containing precursor to cyanocobalamin, is used as an antidote for cyanide poisoning (Antidotes in Depth: A41).

Medicinally, cobalt chloride was combined with iron salts and marketed in the 1950s as “Roncovite”—for the treatment of anemia due to its ability to stimulate erythropoiesis. As recently as 1976, physicians still used cobalt salts to reduce transfusion requirements in anemic patients, despite well-known adverse effects.⁴³ The other common medical use of cobalt is as a radioactive isotope, cobalt-60 (⁶⁰ Co). This γ emitter was formerly used in the radiotherapy of cancers but has been largely replaced by linear accelerators in the Western world. Radiotherapy machines may be targeted by terrorist groups as a source of radioactive material.

Epidemics of cardiomyopathy and goiter termed “beer drinker’s cardiomyopathy”¹⁶ and “cobalt-induced goiter”⁶⁹ occurred between the 1960s and the 1970s. During this period, cobalt sulfate was added to beer as a foam stabilizer. In the 1970s, these epidemics were halted with the discontinued use of cobalt sulfate for this purely esthetic purpose.¹⁰⁶

Current sources of cobalt exposure include chemistry kits,⁷⁴ weather indicators,⁷⁴ antiquated anemia therapies,⁷⁴ cement,⁸¹ fly ash,⁸¹ dyes,⁵³ mineral wool,⁸¹ asbestos,⁸¹ molds for ceramic tiles,⁴⁹ the production of Widia-steel (utilized in the wood industry),¹⁴² mining,⁷⁵ porcelain paint,¹²⁶ orthopedic implants,⁷³ and dental hardware.⁷ A recent area of attention is “arthroprosthetic cobaltism,”^{97,115,130,156,157} which results from cobalt-containing orthopedic implants.¹⁵⁸

The most significant source, however, arises through the formation of cemented tungsten carbide, a “hard metal.” In tungsten carbide factories, powdered cobalt and tungsten are combined through an intense sintering process that exposes the metals to hydrogen that has been heated to 1832°F (1000°C). The first published investigation of these factories reported a 10-fold increase in workspace cobalt concentrations compared to atmospheric concentrations.⁴⁶ These respiratory exposures result in pulmonary toxicity, known as “hard metal disease.” As a result of the original report, occupational studies and preventive measures have greatly reduced the acceptable cobalt exposure concentration in the workplace.

Like other metals, cobalt occurs in elemental, inorganic, and organic forms. The clinical effects of each form are less well defined than more common transition metals such as lead, mercury, or arsenic. Elemental cobalt (Co⁰) toxicity is reported through both inhalational¹⁵⁵ and oral exposures.^68,69 Inorganic cobalt salts typically occur in one of two oxidation states: cobaltous (Co²⁺) or cobaltic (Co³⁺). Inorganic cobalt salts, such as cobaltous chloride (CoCl₂) and cobaltous sulfate (CoSO₄), were historically used for the treatment of anemias^{17,58,64,125,166} and were implicated in the “beer drinker’s cardiomyopathy” epidemics, respectively.^2,101,108

Organic cobalt exposure results from cyanocobalamin (vitamin B₁₂) ingestion, but due to its limited oral absorption and its rapid renal elimination it is considered of low toxicity.¹²² Comparing organic and inorganic forms of cobalt in animal models, cobalt salts are more toxic than the organic forms such as cobalt stearate.²¹

Based on animal studies, oral absorption of cobalt oxides, salts, and metal is highly variable with a reported bioavailability of 5% to 45%.^82,93 In human studies, both iron deficiency and iron overload (hemochromatosis) enhance radiolabeled ⁵⁷CoCl₂ absorption in the small bowel.¹¹⁴ Similarly, inhaled cobalt oxide is about 30% bioavailable,⁹³ but the volume of distribution and elimination half-life are not well defined.

The distribution of cobalt is influenced by plasma proteins, mainly albumin^22,23 and is the basis of the recently Food and Drug Administration (FDA)–approved albumin cobalt binding test (ACB test) for myocardial ischemia. Transferrin, another plasma protein, which normally binds iron can also bind cobalt and may help explain its erythropoietic effect.⁷⁶ Following distribution, cellular uptake is mediated by P2X7 transporter¹⁶⁰ and the active divalent metal transporter 1 (DMT1).²³

Most (50%–88%) absorbed cobalt (organic and inorganic) is eliminated renally, and the remainder is eliminated in the feces.¹⁴⁰ Acutely, an increase in inorganic cobalt concentrations will result in an increased renal elimination.⁷⁹ However, this initial increased elimination decreases rapidly in spite of a large body burden.^3,114

The elimination of cobalt correlates with the patterns of occupational exposure.¹⁵⁵ For example, a worker with a standard work week will have much higher urine cobalt concentrations on Friday morning compared to Monday morning.¹⁵⁵ However, Monday afternoon urine cobalt concentrations may be higher than Friday morning concentrations, which is due to the rapid elimination that occurs following an initial exposure.^3,155 Based on these findings, the exposure over time must be considered when interpreting urinary cobalt concentrations in occupationally exposed individuals.¹²

Like other transition metals, cobalt is a multiorgan toxin. Co²⁺ inhibits several key enzyme systems and interferes with the initiation of protein synthesis.¹³ Polynucleotide phosphorylase, an essential enzyme in RNA synthesis, requires Mg²⁺ to function normally. In vitro, this enzyme only functions at 50% that of normal in the presence of Co²⁺.¹³ It is hypothesized that Co²⁺ is capable of displacing Mg²⁺ from its cofactor site of this enzyme.¹³

Cobalt salts interfere with the Krebs cycle mitochondrial enzyme, α-ketoglutarate dehydrogenase. Due to this inhibition, Co²⁺ increases the rate of anaerobic glycolysis and at the same time decreases oxygen consumption³¹ and thus inhibits aerobic metabolism. In vitro studies demonstrate that other divalent cations, Zn²⁺, Cd²⁺, Cu²⁺, and Ni²⁺, also inhibit α-ketoglutarate dehydrogenase (Chap. 13).¹⁶³ When compared to these other divalent cations, Co²⁺ is not considered a potent inhibitor.¹⁶³ However, this in vitro model demonstrates that Co²⁺ is capable of almost entirely inhibiting the reaction when nicotinamide adenine dinucleotide (NADH) is added.¹⁶³ This study suggests that NADH abundance, such as occurs with chronic ethanol use, may potentiate the inhibition of α-ketoglutarate dehydrogenase.¹⁶⁴

Moreover, cobalt salts are capable of inhibiting α lipoic acid and dihydrolipoic acid by complexing with its sulfhydryl groups.^34,163 These reactions result in the inability to convert both pyruvate into acetyl-CoA and α-ketoglutarate into succinyl-CoA (Chap. 13 and Antidotes in Depth: A24). These combined interactions may explain why chronic ethanol use and cobalt exposure result in cardiomyopathy (see Clinical Manifestations, Cardiovascular).¹⁶³

In addition to enzyme inhibition, CoCl₂ induces oxidant-mediated pulmonary toxicity (see Clinical Manifestations, Pulmonary)¹¹² and neurotoxicity.²³ Xenobiotics implicated in free radical–mediated pulmonary injury are capable of accepting an electron from a reductant and subsequently transferring the electron to oxygen, forming a superoxide free radical (Chap. 12). Cobalt is then capable of accepting another electron, which starts the cycle over again, a process known as redox cycling. This leads to an accumulation of free radicals in the lung due to the abundance of oxygen ready to receive electrons and results in injury.

Within the endocrine system, CoCl₂ is capable of inhibiting tyrosine iodinase.⁸⁵ This enzyme is responsible for combining iodine (I₂) with tyrosine to form monoiodotyrosine and serves as the first step in the synthesis of thyroid hormone (Chap. 56). Inhibition of tyrosine iodinase leads in a decrease of circulating T₃ and T₄ which may result in hypothyroidism (see Clinical Manifestations, Endocrine).

The hematopoietic system is also affected by cobalt salts. Multiple animal models demonstrate that CoCl₂ administration results in reticulocytosis, polycythemia, and erythropoiesis.^{52,83,105,116,117,166} These events occur in both the bone marrow and extramedullary locations.^17,55 Although the pathogenesis remains largely unknown,³⁴ one theory is that cobalt ion binds to iron binding sites such as transferrin,⁷⁶ resulting in impaired oxygen transport to renal cells, which in turn induces erythropoietin production. A second theory in which cobalt is thought to stimulate erythropoiesis is through improving iron availability. In an animal model of anemia, a greater degree of gastrointestinal iron uptake occurs in cobalt-treated rats compared to rats with either hypoxia or nephrectomy.⁵⁵ A similar study of injected cobalt chloride in mice suggested that the increase in iron uptake exceeded that following exogenous erythropoietin.⁴

Finally, within the nervous system, CoCl₂ inhibits neuromuscular transmission by competing with Ca²⁺, another divalent ion. Co²⁺ is 20 times more potent than magnesium with regard to its ability to compete with calcium for a site on the motor nerve terminal.¹⁶² Additionally, the formation of free radicals is theorized as a cause of cobalt associated neurotoxicity.²³

The single acute minimal toxic dose of cobalt salts is not well defined. In fact, varying effects have occurred at variable doses in different patients. Patients with “beer drinker’s cardiomyopathy” received an average daily dose of 6 to 8 mg of CoSO₄ (over weeks to months) and developed toxic effects of acidemia, cardiomyopathy, shock, and death,^80,108 whereas infants being treated for anemia who received much higher daily cobalt doses of an iron-cobalt preparation (40 mg of CoCl₂ and 75 mg of FeSO₄) for 3 months did not develop toxicity.¹³¹ The inconsistency of these findings suggests that multiple factors are responsible for the development of the clinical manifestations of toxicity; in this case the role of ethanol metabolism may be an important variable.⁷⁴

Organ systems affected by acute cobalt poisoning include endocrine,⁶⁸ gastrointestinal,^43,58 central^58,139 and peripheral nervous system,¹³⁹ hematologic,^{52,83,105,166} cardiovascular,⁶⁹ and metabolic.⁶⁹ Chronic inhalational exposures affect the pulmonary system^{26,32,46,87,88,128,147} and dermatologic system.^50,134,165 Radioactive ⁶⁰Co used for radiation therapy is associated with radiation burns (Chap. 134). Unlike acute toxicity, chronic cobalt exposure is not associated with an increased mortality; a cohort study evaluating more than 1100 persons with pulmonary exposures to cobalt salts and oxides over a 30-year period was unable to show an increased mortality rate.¹¹⁰

Acute Exposure

Cardiovascular

“Beer drinker’s cardiomyopathy.”

In 1966, a Veterans Affairs (VA) Hospital in Nebraska cared for 28 white men with a history of beer drinking who presented with tachycardia, dyspnea, metabolic acidosiswith elevated lactate concentration, and congestive heart failure.¹⁰¹ The mortality rate for these cases was 38%, and death occurred rapidly within 72 hours of presentation, due to severe acute metabolic acidosis and cardiac failure.¹⁰¹ Of the survivors, most responded immediately to supportive care and thiamine supplementation and a lack of response was found to be secondary to complications—most commonly, symptomatic pericardial effusions or embolic events.¹⁰¹ Epidemiologic evaluation of this case series revealed that these men habitually drank large quantities of beer.

Ultimately, 64 cases and 30 fatalities were reported from Nebraska.¹⁵⁴ Of the reported 30 deaths, 26 decedents received autopsies. Common postmortem findings were dilated cardiomyopathy and cellular degeneration with vacuolization and edema with a lack of inflammation or fibrosis.¹⁰¹ When cobalt was implicated in the pathogenesis of these deaths, preserved cardiac tissue of eight decedents revealed cobalt concentrations ten times greater than that of controls.¹⁵⁴

Within a year of the Nebraska cases, reports began to emerge from Quebec.¹⁰⁸ Forty-eight beer drinkers (only two of whom were women) developed unexplained cardiomyopathy with a mortality rate of 46%.¹⁰⁸ The only association between these patients was the common consumption of “brand XXX beer”.¹⁰⁸ The producers of this beer had factories in ­Quebec City and Montreal. The only difference between the Nebraska brewery and the one in Quebec was that the latter added ten times the amount of CoSO₄ to the beer as a foam stabilizer.¹⁰⁶ Clinical findings in these cases included tachycardia, tachypnea, polycythemia, and low voltage electrocardiograms (ECGs).¹⁰⁷ Cases began to appear one month after beer with the higher amount of CoSO₄ was released onto the market, and no new cases were reported in Quebec after this beer with more CoSO₄ was removed from the market.¹⁰⁶

In 1972, 20 additional cases occurred in Minneapolis with similar findings of tachycardia, dyspnea, pericardial effusion, polycythemia, and metabolic acidosis with elevated lactate concentration, and there was a mortality rate of 18% acutely and 43% over a 3-year period.⁸⁰ Similar to the previous cases, all were associated with the addition of CoSO₄ as a foam stabilizer in beer.

Since the clinical findings resemble the cardiomyopathy associated with chronic alcoholism⁴⁸ and infantile malnutrition,¹²³ a debate persists as to whether cobalt is the sole cause of this syndrome. Cardiomyopathies due to poor protein intake, vitamin deficiency, and cobalt toxicity have similar histological findings. For example, myocardial biopsy of dogs with cobalt-induced cardiac failure revealed diffuse cytosolic vacuolization, loss of cross striations, and interstitial edema,¹³⁷ all of which are similar to findings of malnutrition.^48,123 However, some other findings may be specific to cobalt-associated cardiomyopathy. For example, a small retrospective analysis revealed myocyte atrophy and myofibril loss to be present in people with cobalt-associated cardiomyopathy significantly more often than in those with idiopathic dilated cardiomyopathy.²⁴

Some animal models of cobalt cardiomyopathy were only able to reproduce pathologic and ECG findings if cobalt and ethanol⁸ were administered, while others required protein deficiency.¹³² Contrary to these studies, several murine and canine models of cobalt poisoning demonstrated cardiac lesions,^62,136,137 cardiac failure^65,136,137 and ECG abnormalities while receiving nutritional supplementation.^63,136

Despite the implication that cobalt-induced cardiomyopathy requires malnutrition or alcoholism, a case of cardiac toxicity following acute cobalt poisoning is reported.^68,69 However, it is difficult to identify other cases reported outside of the aforementioned small epidemics in beer drinkers. In a controlled study of occupationally exposed subjects evaluated with echocardiograms, significantly more cobalt exposed workers had diastolic dysfunction when compared to controls.⁹¹ However, none of these subjects under study developed congestive heart failure.⁹¹ Of the few reported cases of cardiomyopathy associated with arthroplastic cobaltism, patients are typically older than those with the “beer drinker’s cardiomyopathy” cohort and their nutritional status and ethanol use are unreported.^115,121 There are rare reports of cardiomyopathy in chronically exposed workers,^15,30,78 which suggests that the cardiomyopathy reported in the “beer drinkers” cohort is multifactorial and not solely due to cobalt.

Another source of criticism of the role of cobalt in the development of cardiomyopathy is the relatively low dose of cobalt needed to induce heart failure in these patients.⁸⁰ In patients receiving 20 to 75 mg/d of CoCl₂ for various red cell dysplasias, there were no reports of heart failure,⁸⁰ whereas the “beer drinker’s cardiomyopathy” group reportedly consumed only 6 to 8 mg/d of CoSO₄ while drinking 24 pints of cobalt-containing beer.^80,106 All patients who developed cardiomyopathies were malnourished, which supports the theory that a multifactorial nutritional deficiency in the presence of excessive cobalt may be necessary for the development of cardiomyopathy.⁸⁰

Endocrine.

Both acute and chronic cobalt exposures are associated with thyroid hyperplasia and goiter. A series of patients with severe sickle cell anemia treated with cobalt salts developed goiter with varying degrees of thyroid dysfunction,^64,84 including clinical hypothyroidism.⁸⁵ In one patient, the goiter was so severe that airway obstruction developed.⁸⁴

Older occupational data suggest that inhalational exposure to cobalt metals, salts, and oxides may result in abnormalities in thyroid function studies.¹⁵⁵ When 82 workers in a cobalt refinery were compared to age- and sex-matched controls, exposed workers had significantly lower T₃ concentrations.¹⁵⁵ However, more recent studies of cobalt-exposed workers in environments that limit exposure do not demonstrate any changes in serum thyroid markers.⁹⁰

Within the previously mentioned beer drinker’s cardiomyopathy cohort, 11 of 14 decedents had abnormal thyroid histology.¹³³ Among them, the most common findings were follicular cell abnormalities and colloid depletion, which did not exist on thyroid analysis from 11 randomly selected autopsies that served as controls.¹³³ Of the patients with arthroprosthetic cobaltism, one patient developed clinical hypothyroidism,¹¹⁵ and another developed subclinical hypothyroidism.¹²¹

Hematologic.

Anemias of the newborn,^28,77,125 erythrocyte hypoplasia,¹⁴¹ red cell aplasia,¹⁶¹ and kidney failure associated anemia⁵⁸ have all been successfully treated with cobalt salts. Patients undergoing CoCl₂ therapy for these diseases had increased hemoglobin,^58,125 hematocrit,^58,125 and red cell counts.⁵⁸ Unfortunately, the effects did not persist after cessation of therapy.^58,125

A published series of Peruvian cobalt miners working in an open pit at 4300 meters (2.7 miles) elevation developed various clinical effects, including headache, dizziness, weakness, mental fatigue, dyspnea, insomnia, tinnitus, anorexia, cyanosis, polycythemia, and conjunctival hyperemia consistent with acute mountain sickness.⁷⁵ When the study group was compared to age-, height-, and weight-matched high-altitude controls, the study group was noted to have higher chronic mountain sickness scores.⁷⁵ The only difference detected was elevated serum cobalt concentrations in the study group.⁷⁵

In addition to effects on red cells, recent work demonstrates transient hemolysis, methemoglobinemia, and methemoglobinuria from subcutaneous CoCl₂ exposure in mice.⁷⁰ These findings may explain reports of dark urine following cobalt exposure in other animal models.^60,149 However, similar human cases have not been reported.

Other.

Gastrointestinal distress following the ingestion of “therapeutic” doses of cobalt salts¹³⁹ and elemental cobalt are reported.⁷⁴ Decreased proprioception, impaired VIII cranial nerve function, and nonspecific peripheral nerve findings are reported with acute oral CoCl₂ exposures.¹³⁹ Patients with arthroprosthetic cobaltism are reported to have seizures, ­cerebellar deficits, retinopathy, hearing loss, cognitive deficits, and peripheral neuropathy; some of which had concomitant elevations of cerebrospinal fluid (CSF) cobalt concentrations.^{23,97,115,156,157}

Chronic Exposure

Pulmonary.

Both asthma and “hard metal disease” are associated with cobalt exposure. Occupational asthma is reported in hard metal workers with a prevalence of 2% to 5%^26,87,88 at exposure concentrations as low as 50 μg/m³.⁸⁸ As is the case with most causes of occupational asthma, cobalt hypersensitivity–induced asthma is most likely immune mediated rather than toxicologic.^29,87,147 Most hard metal workers are exposed to other metals, such as tungsten (W) and nickel (Ni), in addition to Co, and these other metals may account for some of the occupational asthma that is attributed to cobalt.^144,146 However, in a small but well-performed study in patients with cobalt-associated asthma, intradermal cobalt chloride (CoCl₂) resulted in a positive wheal response in all subjects, and 50% of patients had a positive radioallergosorbent test (RAST) scores, which correlated to the wheal size¹⁴⁵ and suggests Co salts can illicit an immune response independent of the other metals.

Cobalt-associated pulmonary toxicity was first noted in tungsten-­carbide workers^46,66 and was subsequently referred to as “hard metal disease.” Exposures result from the process by which tungsten-carbide is sintered with cobalt. Signs and symptoms of hard metal disease may include upper respiratory tract irritation, exertional dyspnea, severe dry cough, wheezing, and interstitial lung disease ranging from alveolitis to progressive fibrosis. The prevalence of hard metal disease is largely unknown. In one study, 11 of 290 (3.8%) exposed workers were diagnosed with interstitial infiltrates on chest radiographs, but only 2 (0.7%) had a decrease in predicted total lung capacity.¹⁵⁰

Certain individuals who are exposed to large doses of hard metals for prolonged periods never develop disease, which suggests that a susceptible population exists. A glutamate substitution for lysine in position 69 of the β unit HLA-DP has a strong association with hard metal disease, similar to chronic beryllium disease.¹²⁴ Clinically, hard metal disease is difficult to distinguish from berylliosis, although an occupational history should be helpful.

Histopathologic findings of hard metal disease include multinucleated giant cells and interstitial pneumonitis with bronchiolitis.⁹ Elevated concentrations of cobalt in lung tissue can be detected,^129,148 even as long as 4 years after exposure.¹²⁹ In patients with interstitial lung disease, bronchioalveolar lavage (BAL) commonly reveals multinucleated giant cells, type II alveolar cells, and alveolar macrophages.³² The finding of multinucleated giant cells from BAL washing is characteristic of hard metal disease.^{29,35,36,102,159}

A cross-sectional study of more than a 1000 tungsten carbide–exposed workers found an increased odds ratio of 2.1 for having a work-related wheeze when exposed to greater than 50 μg/m³ of Co.¹⁵¹ In the same study, workers with exposures recorded at greater than 100 μg/m³ had higher odds (OR 5.0) of having a chest radiograph profusion score of greater than or equal to 1/0.¹⁵¹ This profusion score, established by the United Nations agency, the International Labor Organization (ILO), is a grading system for pneumoconioses. When used to grade radiographs of asbestosis, this score correlates strongly with mortality risk,¹⁰⁰ reduced diffusing capacity, and decreased ventilatory capacity.^67,111 A score of 0/1 is suggestive but not diagnostic (“negative”), and a score of 1/0 is presumptively diagnostic but not unequivocal (“positive”).⁵ Additional studies have similarly concluded that pulmonary disease occurs when individuals are exposed to concentrations of cobalt that approach 100 μg/m³.⁸⁶ Thus, with a margin of safety, the current threshold limit value is less than 50 μg/m³.

Until 1984, all reported cases of hard metal disease were associated with the combination of cobalt and other metals, such as nickel, cadmium and tungsten.^9,66,88,151 When diamond polishers initiated the use of high-speed grinding disks coated with abrasive microdiamonds embedded in a matrix of cobalt powder, case reports ensued demonstrating similar clinical⁸⁹ and pathologic findings to hard metal disease, strengthening the association with cobalt.^29,35,113 Some authors still contend that the ­presence of other metals^9,66,88,151 and diamond dust^35,59,113 are confounding factors.¹⁵⁵ Like hard metal disease, most reported cases show resolution of symptoms upon removal from the exposure,³⁵ although this is not always sufficient.¹¹³

There are very few reports of isolated cobalt exposures. In an age-and sex-matched study of 82 workers with respiratory exposures to cobalt oxides, cobalt salts, cobalt metal, and no other metal, researchers were unable to detect a difference between exposed (mean of 8 years, time weighted average = 125 μg/m³, 25% >500 μg/m³) and unexposed workers with any objective measured pulmonary tests.¹⁵⁵ Neither group had any abnormality in chest radiography that would suggest pulmonary fibrosis.¹⁵⁵ The only significant pulmonary differences detected were a higher reported rate of dyspnea both on exertion and at rest and the presence of wheezing in the exposed group.¹⁵⁵ These authors concluded that cobalt contributes to the development of pulmonary disease but is not independently responsible for the development of pulmonary fibrosis.¹⁵⁵

Despite the progressive and debilitating nature of hard metal disease, most signs and symptoms improve with cessation of exposure.^98,103,168 Moreover, the length and dose of exposure do not appear to correlate with the presence or severity of illness suggesting that individual susceptibility is the key determinant for developing illness.^98,135

Arthroprosthetic Cobaltism.

It has been known for some time that metal-on-metal alloy orthopedic implants result in elevation of the associated metals in blood, urine, and hair.²⁷ Serum concentrations of cobalt become elevated 3 weeks after surgery and remain elevated through a 5-year study period, which contradicts earlier theories of elevated concentrations being related to the life of the implant.²⁰ Despite the elevation of both chromium and cobalt in measured samples, it appears that the constellation of findings is more consistent with cobalt rather than chromium.¹ However, chromium’s toxicity may present in a more delayed fashioned relative to cobalt.¹ Reported cases of toxicity can follow revisions, dislocations, or first-time arthroplasties of cobalt-containing prosthetics.^97,156,157 Since these findings have been reported, the association of having a cobalt-containing prosthetic, an elevated marker of cobalt burden, and findings of end-organ toxicity has been coined “arthroprosthetic cobaltism.”

Of the reported cases, some of arthroplasties have had excessive wear¹¹⁵ or have been placed following a ceramic arthroplasty resulting in metallosis.¹⁰⁰ In one case, a patient developed hypothyroidism, seizures, peripheral neuropathy, and cardiomyopathy.¹¹⁵ Unfortunately, the diagnosis was not established initially, and the hip revision resulted in improvement of serum cobalt concentrations and thyroid markers but the patient had permanent neurologic sequelae.¹¹⁵ Furthermore, one patient with cobaltism with evidence of cardiac tamponade, hypothyroidism, hearing loss, and polycythemia underwent implant removal and chelation with 2,3-dimercaptoproprane-1-sulfonate, which resulted in decreased blood cobalt concentrations but permanent hearing loss.¹²¹ However, other case reports of metallosis from cobalt arthroplasties suggest clinical manifestations of cobaltism improve following revision.^156,157 It appears that ­lumbar metal-on-metal total disk replacements are associated with smaller elevations in serum cobalt concentrations when compared to hip resurfacing or total hip arthroplasties,¹⁹ but there are no reported cases of clinical cobaltism with lumbar implants. Finally, within this cohort, retinal effects have been noted in at least two case reports.^10,11 The pathophysiology of these cases is poorly understood and confounded by comorbidities typical in this cohort of patients.

Renal.

A single report associates reversible acute kidney injury with the chronic administration of CoCl₂ as treatment for anemia.¹³⁹ Some animal models of cobalt cardiomyopathy demonstrate cellular changes in renal tissue.⁶¹ However, when 26 exposed hard metal workers were evaluated for urinary albumin, retinol binding protein, β₂-microglobulin, and tubular brush border antigens no detectable differences could be found between the study group and controls.⁵⁴ Additionally, patients with cobalt-containing total hip arthroplasties were followed over 2 years with elevated serum cobalt concentrations without any significant effect on serum creatinine over this study period.¹⁴ Based on these few reports, it appears that acute and chronic exposure to cobalt has little effect on the kidneys.

Dermatology.

In a study of 1782 construction workers, 23.6% developed dermatitis and 11.2% developed oil acne while using cobalt-containing cement, fly ash, or asbestos.⁸¹ As in hard metal disease, it is difficult to isolate cobalt as the sole contributor to the development of dermatitis. Nickel (Chap. 99) is the classic toxicant causing dermatitis. Commonly found in some of these construction occupations, it may be implicated in the development of cutaneous manifestations.^50,134 However, consumer products such as piercings, tattoo ink, plastics, clothing dye, makeup, and dental treatments are implicated in cobalt-induced allergic contact dermatitis.⁵³ Furthermore, in vitro studies demonstrate that cobalt can induce production of both Th1- and Th2-type cytokines in peripheral blood mononuclear cells—the same pathway implicated in nickel contact dermatitis.¹⁰⁴ Thus, it appears that cobalt can independently induce a hypersensitivity reaction independent of the presence of nickel.

Reproduction.

In pregnant rats, CoCl₂ exposure neither results in teratogenicity nor fetal toxicity.¹¹⁹ In a case report of a pregnant woman with hard metal disease, the woman was able to bring the pregnancy to term and deliver without complication.¹²⁷ In another case report, a woman with a cobalt-containing hip arthroplasty had repeated joint aspirations, dislocations, and revisions before and during her pregnancy.⁵⁶ Cobalt concentrations of this mother were 138 to 143 μg/L throughout the pregnancy, cord blood concentration was 75 μg/L, and infant blood concentration at 8 weeks was 13 μg/L. Despite these elevated concentrations, there was no clinical evidence of toxicity.⁵⁶ It is hypothesized that only exposures toxic to the mother result in fetal toxicity.⁴²

In mice, chronic exposure to cobalt results in impaired spermatogenesis and decreased fertility without affecting follicular stimulating hormone or leuteinizing hormone, whereas acute exposures did not demonstrate similar reproductive effects.¹²⁰ Additional murine studies discuss the possible interactions between cobalt with iron and zinc, which are both essential elements for spermatogenesis.⁶ Despite these findings, there are no reported human cases that associate cobalt exposure with teratogenicity or impaired fertility.

Carcinogenesis.

Animal experiments with injection of CoCl₂ into soft tissue resulted in soft tissue sarcomas, leading the International Agency for Research on Cancer (IARC) to consider cobalt metal without tungsten carbide and cobalt sulfate and other soluble cobalt (II) salts possibly carcinogenic to humans (group 2B).^{18,33,44, 152,167} It was suggested that cobalt metal associated with tungsten carbide was probably carcinogenic to humans (group 2A).¹⁶⁷ There are case reports and cohort studies that suggest that pulmonary exposure to Co²⁺ increases the risk of lung cancer. However, these studies were unable to control for other known carcinogens such as arsenic.³³ In the largest cohort study to date, which followed more than 1100 workers for more than 38 years, there was no increase in the prevalence of lung cancer.¹⁰⁹

Body fluid cobalt concentrations are not readily available and therefore cannot be used to direct emergent clinical care. Some adjunctive testing that may support a clinical diagnosis of cobalt toxicity should include a complete blood count (CBC), reticulocyte count, erythropoietin (EPO) concentration, and thyroid-stimulating hormone (TSH) concentration. The results of these tests may reflect the level of exposure or potential toxic manifestations.

Cardiac Studies

ECG, echocardiogram, and radionuclide angiocardiography with ⁹⁹Tc(RNA) are useful screening tests for detecting abnormalities associated with cobalt cardiomyopathy and/or pulmonary hypertension due to hard metal disease.³⁰ It is important to remember that these cardiac tests are neither specific nor diagnostic of cobalt-induced cardiomyopathy.

Pulmonary Testing

Patients with hard metal lung disease may demonstrate bilateral upper lobe interstitial lung disease on chest radiograph. However, patients may have signs and symptoms of disease without specific radiographic findings.¹²⁷ Pulmonary function testing in occupationally exposed workers may show decreased vital capacity¹²⁷ and a decrease in transfer factor for carbon monoxide, both of which may be useful in identifying patients at risk for developing pulmonary fibrosis.¹⁵⁴ Some authors suggest an inversion of CD4/CD8 ratio in BAL washings as a useful tool for diagnosis and evaluation of progression of illness and that normalization is a marker for improvement.¹²⁸ Despite these available tests, a definitive diagnosis of hard metal disease requires a tissue sample with findings of multinucleated giant cells in the setting of interstitial pulmonary fibrosis.

Cobalt Testing

Cobalt, as stated above, is primarily eliminated in the urine and to a lesser extent in the feces making urine cobalt evaluation most appropriate.⁹³ The difficulty lies in the interpretation of the result. Cobalt is detectable in the urine after inhalational exposure and reflects elimination kinetics that are rapid during an initial exposure but slow after prolonged exposure.^93,138 Due to this variable elimination pattern, it is difficult to interpret both urine and blood concentrations unless the dose and length of exposure are precisely known. Furthermore, the defined patterns may be applicable only to shift workers utilizing soluble forms of cobalt.⁹³

Further complicating the interpretation of urinary cobalt concentrations is the abundance of organic cobalt in the form of vitamin B₁₂. A detailed vitamin supplementation history is required prior to the interpretation of a urine or blood cobalt concentration, as a diet regimen high in vitamin B₁₂ may increase urine cobalt concentrations. For this reason, speciation of cobalt has been investigated. The ratio of inorganic to organic cobalt is higher in occupationally exposed workers (2.3) when compared to controls (1.01), independent of the wide variations of urinary cobalt concentrations.⁵⁷ This is a promising area of study for the evaluation of cobalt exposed workers.

Toxic concentrations of cobalt in serum and urine are poorly defined. Published literature on “normal values” is fraught with variability, which may reflect differences in the population under study techniques for measurement. Normal serum concentrations of cobalt are frequently reported as 0.1 to 1.2 μg/L.^{3,16,68,69,72,143} In comparison, a single acutely poisoned patient had a reported serum concentration of 41 μg/L.⁶⁸ In the “arthropro­sthetic cobaltism” group, serum cobalt concentrations ranged from 24 to 625 μg/L.^{56,98,116,122,157} In contrast, several case series of metal-on-metal hip arthroplasties and resurfacings demonstrate peak serum cobalt concentra­tion not to exceed that of 11.5 μg/L barring one outlier of 35 μg/L at 1 year postprocedure.^14,19,20 Based on this data, some recommend that the risk of “arthroprosthetic cobaltism” be suspected when serum cobalt concentrations exceed 10 μg/L.¹⁵⁷

Normal reference urine cobalt concentrations are between 0.1 and 2.2 μg/L.^{3,16,68,69,72,118,143} In contrast, an acute elemental cobalt ingestion resulted in a concentration of 1700 μg/L on a spot urinalysis several days after the exposure.⁶⁹ In the “arthroprosthetic cobaltism” cases, one patient had a urine cobalt concentration of 16,500 μg/L.¹¹⁵

Chronic exposures should be evaluated differently as discussed above (see Toxicokinetics). Exposed workers, without clinical disease, have reported spot urine concentrations that range from 10 μg/L to several hundred μg/L.⁷¹

Acute Management

Patients with acute cobalt poisoning require aggressive therapy. It is reasonable to conclude that the same decontamination principles used for other metals apply to cobalt. There have been no studies examining the benefit of gastric emptying, activated charcoal, or whole bowel irrigation. An attempt of whole-bowel irrigation for radiopaque solid forms of cobalt should be made prior to endoscopic or surgical removal. Regardless of the decontamination procedure utilized, chelation therapy should be not be ­initiated until the gastrointestinal cobalt source has been removed. If there is a large stomach burden in solid form, endoscopic or surgical removal may be of ­benefit,⁶⁹ keeping in mind that the administration of activated charcoal prior to surgical removal may obscure visualization of the surgical field (Chap. 8). After decontamination, reduction of tissue burden and prevention of end-organ toxicity is the next crucial step.

The data on chelation therapy is unfortunately limited to animal models^{37,38,39,40,41, and 42,94–96} and human case reports.^69,121 The basis of chelation therapy originates from the mid 1900s when oral protein intake was found to result in a reduction of cobalt toxicity in calves⁴⁵ and rats.⁶² Certain sulfur containing proteins serve as good chelators for some transition metals.

In a series of animal models of acute cobalt toxicity, several agents N-acetylcysteine (NAC), succimer, ethylenediaminetetraacetic acid (EDTA), glutathione, and diethylenetriaminopentaacetic acid (DTPA) were evaluated for their ability to enhance urinary and fecal elimination of cobalt.⁹⁶ Succimer and EDTA were able to enhance fecal elimination.⁹⁶ Glutathione and DTPA were able to enhance urinary elimination, and NAC was able to enhance elimination by both routes.⁹⁶

In two separate animal studies, NAC reduced tissue burden and injury due to cobalt in the liver and spleen.^41,96 Glutathione, another sulfur-containing protein, also reduced tissue concentrations of cobalt but only in the spleen.⁹⁶

The sulfur-containing proteins, NAC,^38,41 L-cysteine,^37,38 L-methionine,³⁸ and L-histidine,⁴² were studied for their ability to reduce mortality in rats that were administered an LD₅₀ of CoCl₂ orally and intraperitoneally. Therapies with NAC, L-cysteine, and L-histidine⁴² are more effective than L-methionine in protecting against mortality, upward of 100% protection. In a similar fashion, Na₂ EDTA effectively reduced mortality.⁴⁰ In all these therapies, the successful reduction in mortality from these therapies is predicated on its early administration.^{37,38,40,41, and 42}

In a murine model, L-cysteine, NAC, glutathione, L-histidine, sodium salicylate, D,L-penicillamine, succimer, N-acetylpenicillamine (NAPA), diethyldithiocarbamate (DDC), BAL, 4,5-dihydroxy-1,3-benzene disulfonic acid, Na₃ CaDTPA, Na₂CaEDTA, and deferoxamine mesylate were each evaluated for exposures to an LD₅₀ and an LD₉₉ of CoCl₂.⁹⁵ Sodium salicylate, NAPA, DDC, BAL, 4,5-dihydroxy-1,3-benzene disulfonic acid, and deferoxamine mesylate were all ineffective at improving survival following an LD₅₀.⁹⁵ Chelators that were seemingly effective at an LD₅₀ and not at an LD₉₉ were L-histidine and D,L penicillamine.⁹⁴ NAC, L-cysteine, and succimer were able to improve survival by 40% to 50%.⁹⁵ The most effective chelators at LD₉₉ were EDTA and DTPA.^94,95 An expanded analysis of these data revealed that EDTA and DTPA had a better therapeutic index when compared to succimer.⁹⁵ In this study, BAL was ineffective as is suggested in an in vitro study where BAL was unable to chelate Co²⁺ that is already bound to α-ketoglutarate.¹⁶³

Human chelation data are available from two case reports.^69,121 In the first, a child ingested multiple elemental cobalt containing magnets yieldinga serum cobalt concentration of 41 μg/L.⁶⁹ Five days of 50 mg/kg/d of intravenous CaNa₂EDTA enhanced renal elimination of cobalt, and the patient’s metabolic acidosis and the cardiac dysfunction also resolved simultaneously.⁶⁹ In the second, a 56 year-old man developed “arthroprostheticcobaltism” with elevated serum (506 μg/L) and CSF concentrations (8.5 μg/L) of cobalt and evidence of hypothyroidism, cardiomyopathy with pericardial effusion, polycythemia, and both central and peripheral neuropathy.¹²¹ Unlike the previous case report, this patient was treated 2,3-dimercaptoproprane-1-sulfonate (DMPS) at a dose of 14 mg/kg for 6 days and followed by 4 mg/kg for 5 days for a total of 10 g of antidote. The authors were able to demonstrate a 26% reduction in serum cobalt concentrations 1 month following treatment.¹²¹ However, despite removal of the prosthesis and chelation therapy, the patient had permanent hearing loss and a persistently elevated serum concentration.¹²¹ This permanent hearing loss and persistently elevated serum concentration are most likely due to a delay of 7 months from onset of symptoms and metallosis of tissue to treatments.

In conclusion, based on a single human case report, several animal studies and safety profiles, CaNa₂EDTA and NAC can be used as antidotal therapy. Indications for treatment should include patients who demonstrate end-organ manifestations of toxicity. This includes metabolic acidosis and cardiac failure. Other manifestations of severe cobalt toxicity such pericardial effusion, clinically significant goiter, and hyperviscosity syndrome should be treated aggressively with pericardiocentesis, airway protection, and phlebotomy, respectively. Based on years of experience with lead, CaNa₂EDTA should be administered as doses of 1000 mg/m²/d by continuous infusion for 5 days. If the diagnosis is confirmed and signs of cardiac failure and metabolic acidosis persist after 5 days, an alternate chelator (succimer or DTPA) can be started. Similarly, NAC dosing should be based on the acetaminophen (APAP) experience. The 20 hour intravenous NAC protocol should be initiated and continued as in the case of fulminant hepatic failure (Antidotes in Depth: A3) for as long as the patient can tolerate therapy or continued if cardiac failure or acidemia persists. If there are contraindications to intravenous NAC, oral NAC can be administered using one of the APAP treatment regimens. Thiamine hydrochloride should be administered to all patients presenting with or without overt cardiomyopathy independent of whether the patient is alcoholic or malnourished. The dose of thiamine is not well defined but should be based on its safety and clinical experience with the treatment of Wernicke encephalopathy. The daily administration of 100 mg of parenteral thiamine can be initiated with increasing doses to 100 mg every hour for life threatening manifestations (cardiac failure and metabolic acidosis) (Antidotes in Depth: A24).

Arthroplasty Cobaltism

It appears that clinical symptoms and serum cobalt reduction respond well to arthroplasty revision^115,121,156 and potentially chelation therapy.¹²¹ The key to treatment of this entity is early clinical suspicion with the constellation of findings of cardiomyopathy, hypothyroidism, polycythemia, and peripheral and central nervous system impairment. Early arthroplastic revision supports the decontamination tenets within this text and published in case reports;^115,121,156 however, emergent revisions may not be practical in an acutely ill patient. Therefore, the only modality of treatment other than supportive may be chelation. However, based on limited literature, it is unclear if chelation in the presence of a large reservoir of metal can mobilize cobalt. However, the single case report of DMPS treatment for a patient with “arthroprosthetic cobaltism” (including cardiomyopathy, neurologic derangements, hypothyroidism, and polycythemia) was able to demonstrate a decrease in serum cobalt concentration and an increase in urinary cobalt concentration postchelation.¹²¹

Occupational/Chronic Exposure

As is always the case of occupational poisonings, prevention is of paramount importance. The utilization of skin and respiratory protection and improvement of personal hygiene has reduced exposure and subsequently the amount of urinary cobalt in occupationally exposed workers.⁹² Barrier and emollient creams cannot prevent the dermatitis associated with cobalt metal exposures.⁵¹

Large statistically significant reductions in urinary cobalt concentrations were demonstrated after the implementation of aspirator systems over machines in the production of Widia-steel.²⁵ These aspirators were found to reduce ambient cobalt concentrations by as much as a factor of six.⁴⁷

• Acute cobalt toxicity results in multiorgan system toxicity, including the cardiac, endocrine, hematopoietic, gastrointestinal, and neurologic systems.

• Arthroprosthetic cobaltism is unique in that there is a large reservoir of metal, and the clinical presentation may be subacute and mistaken for other disease processes.

• The evaluation and treatment of all patients will be determined by the cobalt source and type (elemental, inorganic, or organic), route of poisoning, and time and duration of exposure.

• Treatment of patients with chronic occupational exposures to cobalt is mainly symptomatic and is often dependent on improved industrial hygiene and removal from the source.

• Acute poisoning with end organ manifestations may require aggressive gastrointestinal decontamination and chelation therapy utilizing CaNa₂EDTA and NAC, and at times, succimer.

References