Cobalt and Cancer: Exposure and Lung Cancer Risk

Cobalt is a hard, silvery metal. Its atomic number is 27 and its CAS number is 7440-48-4. It is mined as a by-product of copper, nickel, zinc, lead and platinum ores. The United States has barely mined it for decades and refined almost none of it in 2012.

US use in 2012 came to about 8,420 metric tons. The breakdown looked like this:

• Superalloys, mostly for jet engines and turbines: 48.0 percent.
• Chemicals and ceramics, including pigments, paint driers and catalysts: 27.3 percent.
• Cemented carbides, the hard cutting-tool material: 9.2 percent.
• Other alloys, including magnetic and wear-resistant grades: 8.3 percent.
• Steels: 6.5 percent.

Worldwide, the single largest use is rechargeable battery electrodes. That includes the lithium-ion cells in phones, laptops and electric cars.

The Report on Carcinogens is the federal list kept by the National Toxicology Program. Cobalt appears on it twice, as two separate entries, and the difference matters.

The first entry covers cobalt and cobalt compounds that release cobalt ions inside the body. These are "reasonably anticipated to be human carcinogens." That is the lower of the two NTP levels. The listing dates from the Fourteenth Report in 2016 and replaced an older, narrower listing of cobalt sulfate. It rests on sufficient evidence in animals plus mechanism data.

The wording is deliberate. NTP concluded that releasing cobalt ions in the body is the key step. Compounds that dissolve in water release ions outside the cell, and the ions then enter through channels in the cell membrane. Poorly soluble cobalt particles take a different path. Cells swallow them whole into lysosomes, the acidic recycling compartments, where the particles dissolve and release ions inside the cell. Either way the cobalt ion is what does the damage.

The second entry covers cobalt-tungsten carbide powders and hard metals. These are also "reasonably anticipated to be human carcinogens," but for a different reason. Here the evidence comes from people.

Hard metal is a composite. Tungsten carbide particles are bound together with powdered cobalt and pressed into a solid at high heat, a process called sintering. Commercial grades run from 50 to 97 percent tungsten carbide and 3 to 16 percent cobalt. About 70 percent of production goes into cutting tools and 30 percent into wear-resistant parts for mining and grinding.

NTP based the listing on three findings across the worker studies:

• Deaths from lung cancer were consistently higher among hard-metal manufacturing workers.
• Risk was higher in workers with higher exposure.
• The exposure-response pattern could not be explained away by tobacco smoking.

NTP also states the limit of that evidence plainly. The data are limited because there are few studies of independent populations.

Laboratory work shows that cobalt-tungsten carbide is more toxic to cells than either cobalt powder or tungsten carbide alone. The leading explanation is a surface reaction between the two particles. Cobalt metal can pull oxygen apart, but slowly. Tungsten carbide does not react with oxygen at all, but it conducts electrons well. When the two sit together, electrons move from the cobalt to the carbide surface. That speeds up oxygen reduction and produces more reactive oxygen species, the unstable molecules that damage DNA.

The same mixture causes a lung disease that cobalt alone does not. It is called hard-metal disease. Under the microscope it shows as giant-cell interstitial pneumonia, and it can progress to lung fibrosis, meaning permanent scarring. That is a separate illness from cancer, and it can appear much sooner.

IARC re-examined cobalt in Volume 131, published in 2023. It did not issue one verdict. It issued six, because different cobalt compounds behave differently.

• Cobalt metal, without tungsten carbide or other alloys: Group 2A, probably carcinogenic to humans.
• Soluble cobalt(II) salts: Group 2A.
• Cobalt(II) oxide: Group 2B, possibly carcinogenic to humans.
• Cobalt(II,III) oxide: Group 3, not classifiable.
• Cobalt(II) sulfide: Group 3.
• Other cobalt(II) compounds: Group 3.

For cobalt metal the human evidence was rated inadequate, the animal evidence sufficient, and the mechanistic evidence strong. That combination is what produces a 2A rather than a 1. It means the case does not rest on human cancer data.

Cobalt-containing alloys are used in joint replacements. Urine testing shows the pattern clearly. The highest cobalt levels turn up in exposed workers and in people whose hip implants have failed. Levels are lower with stable implants, and lowest in the general public. For comparison, US general-population urinary cobalt averaged between 0.316 and 0.391 micrograms per liter in national survey data from 1999 and 2014.

On cancer risk from implants, NTP does not claim to know. It writes that it is not possible to work out the quantitative cancer risk from cobalt ions released by surgical implants, because both the animal studies of cobalt alloy implants and the studies of patients have real limitations. Elevated cobalt in blood or urine after a failed implant is a recognized medical problem for other reasons, and it is managed by the surgeon who placed the device.

Cobalt sits at the center of the vitamin B12 molecule, and B12 is essential. NTP wrote an explicit exclusion into the listing. Vitamin B12 does not meet the criteria, because it does not release cobalt ions. It travels through the body intact, bound to specific carrier proteins. Food cobalt and B12 supplements are not what these listings are about.

• OSHA permissible exposure limit: 0.1 mg/m3 for cobalt metal, dust and fume, as an 8-hour average. NTP notes this limit was adopted from a 1968 guideline soon after OSHA was created, and warns it "may not reflect the most recent scientific evidence and may not adequately protect worker health."
• NIOSH recommended exposure limit: 0.05 mg/m3.
• ACGIH threshold limit value: 0.02 mg/m3 for cobalt and inorganic compounds. This is five times below the OSHA rule.
• ACGIH biological exposure index: 15 micrograms of cobalt per liter of urine. Unlike an air limit, this measures what actually got into the worker.

The jobs with real exposure are specific: hard-metal production, tool sharpening and grinding, cobalt refining, diamond polishing, and some battery and pigment work. Grinding sintered parts is a high-risk task, because it makes fine respirable dust from a solid object.

If this is your trade, two things are worth doing. Get the biological monitoring, not just the air sampling, since urine testing reflects your own absorbed dose. And describe the work in detail to any doctor evaluating a cough or breathlessness, because hard-metal disease is missed when nobody asks. Our page on lung cancer covers how lung symptoms are worked up, and carcinogens at work explains how job exposures are grouped and recorded.

• IARC Monographs volume 131: cobalt
• IARC volume 131: summary of final evaluations
• NTP Report on Carcinogens: cobalt
• OSHA table Z 1