Carcinogenicity of mercury and mercury compounds

BOFFETIA P, MERLER E, VAINIO H. Carcinogenicity of mercury and mercury compounds. Scand J Work Environ Health 1993;19:1-7. Mercury and mercury compounds are widely used in modern society, but only sparse data are available on their carcinogenicity. Methylmercury chloride causes kidney tumors in male mice. Mercury chloride has shown some carcinogenic activity in male rats, but the evidence for female rats and male mice is equivocal. Other mercury compounds and metallic mercury have not been tested adequately in experimental animals. Epidemiologic data are available for chloralkali workers, dentists and dental nurses, and nuclear weapons workers, three groups occu pationally exposed to low levels of mercury and its compounds, but those highly exposed in the past, such as miners, or populations which have suffered massive environmental exposure have not been adequately studied. However, the sparse epidemiologic data point toward the possibility of a risk of lung, kidney, and central nervous system tumors. Better data are needed on the carcinogenicity of mercury and mercury compounds in humans and experimental animals.

T wo types of mercury mines exi st. There ar e ores co ntaining nati ve mercu ry (free metal in minute droplet s or qui ck silver), to which miner s are exposed during quarrying; such ore is found in the mines of Idrija in Slovenia, A lm aden in Spain, and Huancauelica in Peru . In other m ines mercury is present as cinnab ar (crys ta lline mercury sulfide) , as in the mine of Monte Amiata in Ital y and in No rth American mines.
The te chnology used in the mines and th e levels of e xposu re ha ve not be en well documented, although data on air level s of mercury have been published for a few mines and furnaces. In Idrija, airborne mercury levels in 1950 were in the range 0.05-5.9 mg · rrr? in the mine and 0.17-1.1 mg· m-3 in the sme lte r (4) . Similar value s were found in 1963 (5).
In seve ra l mines, reduc ed sh ifts and the rot ation of workers between m ining and roasting are measures th at have been taken to mitigate the harmful eff ect a Based on references 1-3. Other producers include Australia, Canada, Chile, Colombia, Finland, Germany, Ireland, Japan , Peru, Philippines, Romania, Tunisia. b 1978 production.
Mercury and its com po unds have been used si nce an ci ent times, but the uses have changed profoundly si nce the industrial revolution as a consequence of the greater availability of the metal and the recognition of the health effects of occupational e xposure a nd the hazard s due to enviro nm enta l pollution and ac cumulation (1).
Th e chemical forms of mercury ca n be classifi ed as in or ganic and or ganic. M ercury assumes three oxidation states; namely, it can be metallic, me rcurou s, and mercuric. Inorganic and organic compounds contain mercury in the latter two oxidation states.
Organic forms are those in which mercury is attach ed cov a lently to at lea st one ca rbo n atom. Methylmerc ury is th e most important in terms of human exposure.

Production
A comparison of es tim ates of the worldwide pr oduction of mercury during th is century s uggests a shift from a stro ng co nc entra tio n in Italy and Spain in th e early part of th e century to a much w ider range of producing countries , including newly industrial ized countrie s such as China and Algeria (ta ble I).

Uses
An estimate of the con sumption of mercury and mercury compounds in the early 1970s in indu strialized countries is as follows: chloralkali industry 25 %, ele ctrical instruments 20 %, pa ints 15%, medical and precision instruments 10%, dent al amalgams 3%, laboratories 2%, other uses 20 % (6) . Th e main industrial uses of mercury are presented in table 2. One of the mo st import ant uses is in the chloralkali industry, wh ere airborne and urinary or blood mercury levels have been monitored for decades. A decrease in exposure over time has been observed (13,14). Factories differ greatly, however, in exposure levels even in the same country (15,16), and exposure varies greatly among jobs (17,18). Ma ssive mercury intoxication occurred in Minamata Ba y, Japan, between 1953 and 1965 , due to the di sch ar ge of large quantities of methylmercury into a river as a result of a side reaction in the synthesis of aceta ldehy de from acetylene ( 19) .
Other major use s include dental ama lgam and organic mercury pesticides. A number of epidemics of mercury poi soning oc curred aft er the consumption of product s deri ved from grain tre ated with eth yl-or methylmercury pesticid es. Th e mo st important intoxication occurred in Iraq in 1971-1 972, and it was du e to the consumption of grain tre ated with methylmerc ur y-based anti fung al agents (20).

Mercury toxicity
Detailed reviews of the met abolism of mercury and meth ylmercury have been publ ished (2 1, 22). Inhaled Table 2. Major industrial uses of mercury (1,7). mercury vapor reaching the alveoli pa sse s nearly completely into the blood stream, whe re it is ox id ized in red blood cells, the liver , and the kidney to the div alent form (ionic mercury) by the hydrogen peroxide catalase complex (21). Deposition and accumul ation occur primarily in the brain and kidney, but also in several other organs (thy roid, pituitary gland, brain , liver, pancreas , testes or ovary, pro state). Th e oxidative ca pac ity of red blood cell s ca n be overwhe lmed by a high rate of entry or inhi bited by ethano l. In suc h ca se s, mercury rem ain s available for transport to the brain. Elimination appear s to be slow er (taki ng even years) from the brain and kidney (23). The central nervous sys tem is the pr incipal target organ of methylmercury, whi ch affects ma inl y spec ific areas of the brain , such as cerebellum and temporal lobes.
Acute and chronic toxic effects of mercury and mercury compounds in high-and low-level doses caused by poisonings or occupational expo sures have been rev iewed extensively (I, 7,21,22 ). Th e critical organs are the central nervous system and the kidne ys.

Inorganic mercury
Th e most ob viou s ge notox ic effect of inorganic mercury compounds is the induction of C-mitosi s with inacti vati on of the mitotic spindle. Thi s phenomenon results in ane uploidy and pol yploidy and is probabl y du e to th e action of the mercury ion on sulfhydryl gro ups in the sp indle apparatus (24)(25)(26) . The ability of inorganic mercury to induce genetic mutations appears, however , to be low (25).
Cy togenetic studies on expo sed hum an s have generall y shown no effects on chromosomes in the peripher al blood lymphocytes (27)(28)(29)(30). In a recent study on 26 chloralkali workers exposed to inorganic Use Production of electricaldevices and batteries Production of industrial and controlinstruments Chemical industry For example, fluorescent lamps, rectifiers, power cells, switches For example, barometers, thermometers, hygrometers Production of acetaldehyde from acetylene; production of Chlorine and caustic soda (chloralkali industry) Mainly replaced by electroplating Important source of environmental contamination (8) Used in fur carroting; has caused many cases of severe intoxication (9,10); has been replaced by hydrogen peroxide or sulfuric acid Used in the lithium isotope separation process (11) Main uses include dental amalgam. treatment of syphilis , antiseptic and diureticdrugs and brain scintigrams ; most uses have been discontinued Used in antifouling and mildew proofing paints for boats, ships and wood houses (12); used in paints for pottery and porcelain Methylmercury and other alkyl mercurials; mainly used for the prevention of seedborne cereal diseases; banned in many countries Used as slime control agent (7); use discontinued in the 1970s mercury, no increase in lymphocyte micronuclei was observed (31). However, a significant correlation was reported between previous exposure to mercury (cumulative exposure or number of blood mercury peaks) and the frequency of lymphocyte micronuclei (31); this finding suggests an accumulation of cytogenetic effects in T-lymphocytes.
In one study, an increase in chromosome aberrations and micronuclei in peripheral lymphocytes was found among workers exposed to mercury fulminate when the workers were compared with referents (32).
In human whole-blood cultures, mercuric chloride has caused a dose-dependent increase in sister chromatid exchanges (33).

Methylmercury
Organic mercury is more toxic than inorganic mercury in cultured mammalian and human cells (34,35). Organic mercurials, including methylmercury, have been shown to be about 10 times more effective than mercury chloride in inducing abnormal mitosis and single-strand breaks in cellular deoxyribonucleic acid (DNA) (35,36).
The dose of methylmercury required to increase the frequency of sister chromatid exchanges in cultures of human whole blood cells in vitro was found to be about one-fifth of that of mercury chloride (33). Methylmercury can cause spindle disturbances in mammalian cells in culture, an effect which appears to be mediated by the interaction of mercury with sulfhydryl ions (37). In Drosophila melanogaster, the induction of nondisjunction (38) and gender-linked recessive lethal mutations has been found after treatment with methylmercury (39).

Metallic mercury
One study has reported the occurrence of local sarcomas in rats after intraperitoneal injections of metallic mercury (40).

Mercury chloride
One large study was carried out by the National Toxicology Program in the United States to investigate the carcinogenicity of mercury chloride in rats and mice at concentrations of 2.5 and 5 mg . kg:' (rats) and 5 and 10 mg· kg:' (mice) (41). Male rats had an increased incidence of papillary hyperplasia and squamous cell papilloma of the forestomach. Two of the 50 female rats in the highest dose group had a squamous cell papilloma of the forestomach (none in the remaining groups). Among the male mice, there were three adenocarcinomas of the renal tubule in the two treated groups [two in one group (N = 49) and one in the other (N = 49)] and no such tumors in the reference group. No tumor increase was found among the female mice. Therefore, mercury chloride Scand J Work Environ Health 1993, vol 19, no I showed some evidence of carcinogenic activity in male rats, equivocal evidence for female rats and male mice, and no evidence for female mice.

Methylmercury chloride
Rats. Only limited data on the carcinogenicity of methylmercury chloride in rats are available. A study on weaning animals reported no increase in tumor incidence at concentrations up to 2.5 mg . kg:" (42). Another study on adult rats also failed to detect any increase in tumors among animals treated with up to 10 mg . kg 3 (43,44). Another study in which methylmercury chloride was given to pregnant rats, together with increasing concentrations of ethylurea, exposure to methylmercury did not affect the incidence of neurological tumors in progeny (45). However, the tumors in the progeny of the methylmercury-treated rats tended to appear earlier than those in the progeny of rats not treated with methylmercury.
Mice. Four studies on the long-term carcinogenicity of methylmercury chloride among mice have been reported (46---49). All of them showed an increased incidence of renal tumors among treated males (the minimal dose that led to a significant increase was 10 mg . kg'), but not among treated females. The possible hormonal dependence of renal tumors induced by methylmercury chloride in male mice was investigated in detail in one of these studies, in which a significant increase in kidney tumors was seen among noncastrated mice treated with methylmercury chloride, but not in castrated treated mice or in referents (48). Another study on female mice found an increase in the incidence of pulmonary adenomas in the group treated with methylmercury chloride and urethane, as compared with animals treated with urethane only (50). In conclusion, methylmercury chloride is clearly carcinogenic in male mice, and a carcinogenic effect in female mice cannot be excluded at present.

Cancer epidemiology studies
Only a few epidemiologic studies have analyzed cancer mortality or incidence in relation to occupational or environmental exposure to mercury. The studied occupations possibly exposed to mercury are miners (51), workers in chloralkali plants (13,52,53), dentists (54-58), workers in the nuclear weapons industry (II), disinfectant applicators (59), and hat makers (60). The design of these studies, as well as their main results, are presented in table 3. Two population-based case-referent studies provided additional information (61, 62). They are summarized in table 4.
Although farmers are probably the largest occupational group with exposure to mercury and mer-3 Scand J Work Environ Health 1993, vol 19, no I Table 3. Epidemiologic studies that investigated specific groups of mercury-exposed workers. (US = United States, SMR = standardized mortality ratio, 95% CI = 95% confidence interval, UK = United Kingdom, SIR =standardized incidence ratio, CNS = central nervous system, PMR = proportionate mortality ratio) cury compounds, no epidemiologic study has analyzed exposure to mercury-containing pesticides separately from exposure to other groups of pesticide s.
Cancer risk has also been investigated among individu als who, in the 1950s, suffered acute mercury poisoning from their intake of pollut ed fish in Minamata, Japan. No over all excess risk of cancer was found in the follow-up of this population, eith er among over 700 poisoned persons or among inhabitants (about 4000) of the polluted area (63-65). However, excesses of cancer of the esophagus , liver, breast, and lung, as well as leukemia (based on small numbers of cases) were detected (64) .

Comments
Higher risk of prostate cancer for substantial exposure No trend with number of fillings lar to the one in Minamata occurred; the victims of a mass poisoning that took place in Iraq in 1971-1972, when seed grain treated with a methylmercury fungicide was used to prepare homemade bread in rural communities (22); and groups of Canadian Indians, who had been exposed seasonally over a long period of time to methylmercury though fish consumption (22). Cancer occurrence in these populations has not been studied.
Finally, a study from a rural area in Poland showed a higher mercury hair content in acute leukemia patients than in referents, but no difference between persons with chronic granulocytic or lymphocytic leukemia and referents (66).

Lun g cancer
A study (51) on miners in the United States suggested that the risk of lung cancer is higher among mercury miners than among other miners, both among silicotics and nonsilicotics. Two prospective studies on chloralkali workers found an overall increased risk of lung cancer (13,53), but no trend according to latency, duration of employment, or estimated cumulative exposure to mercury. The data on hat makers support an association between mercury exposure and lung cancer (60). A population-based study from Montreal found an increased risk of lung cancer for any exposure to mercury, but not for substantial exposure (61). Studies on dentists have not found an exce ss risk of lung cancer (54-58).
Mercury miners can also be exposed to other known or suspected lung carcinogens, such as radon and silica, but no apparent confounder has been identified for chloralkali workers. Potential confounders such as smoking and social class cannot be ruled out however. They might have either increased the risk of lung cancer among miners and chemical workers or decreased it among dentists. The consistency of the findings on lung cancer in different industries suggests a true carcinogenic effect of mercury on the human lung. However, this suggestion is limited by the lack of a dose-response relationship in analyses by latency, duration of employment, or estimated cumulative mercury exposure.

Kidney cancer
On the basis of the evidence from human toxicology and animal carcinogenicity studies, the kidney is among the most likely target organs of a carcinogenic effect of mercury. No evidence of an increased risk of kidney cancer can be derived from the studies on chloralkali workers (13,53), and only one study on dentists (57) and a study on nuclear weapons workers (I I) reported small increases in risk, on the basis of four deaths in each study. Therefore, the data available do not suggest a strong carcinogenic effect of mercury on the human kidney, but they are not sufficient to exclude it either.

Tumors of the central nervous system
Great concern about the carcinogenicity of mercury on the central nervous system has been raised by the results of a study on Swedish dentists and dental nurses (54). Although some studies have suggested a similar effect among other groups of dentists (55, 57) and among nuclear weapons workers (II), the studies on chloralkali workers (13, 53), seed disinfectant applicators (59), and a population-based study from Australia (62) do not support this association. Possible explanations for the inconsistency of these results are the different effects of metallic mercury as compared with organic or inorganic species and the existence of other risk factors for tumors of the central nervous system among dentists.

Tumors in other organs
Three studies on dentists found an increased risk of pancreas cancer (55, 56, 58), and two studies described an increased risk of Iymphohematopoietic neoplasms (55, 58). Although the consistency of the results on dentists suggests the presence of a risk factor, the lack of positive findings in other groups exposed to mercury does not support the hypothesis of a role of mercury.
Increased risks of tumors from other sites have been sporadically reported among groups exposed to mercury, for example, prostate cancer among individuals exposed to mercury in the study from Mon-S Scand J Work Environ Health 1993, vol 19,no I treal (61). Although these results may have occurred by chance, they should be kept in mind in future investigations.

Concluding remarks
Occupational and environmental exposure to mercury and mercury compounds occurs in most countries. However, data on the possible carcinogenicity of mercury and its compounds are sparse. The few studies available strongly suggest a genetic activity of mercury compounds in human and other species. Data on long-term animal experiments on methylmercury exposure strongly suggest a carcinogenic response in the kidneys of male mice. Data for other species and other sites are of insufficient quality to exclude other carcinogenic effects. Among other mercury compounds, only mercury chloride has been extensively studied in long-term experiments, and increased tumors were found in the kidney and forestomach of male rats and, to a less extent, female rats and male mice. Metallic mercury has not been adequately tested in experimental animals.
Epidemiologic results are not sufficient, at present, to allow a conclusion to be drawn about the carcinogenicity of mercury and its compounds in humans. Some of the populations that have been studied (chloralkali workers, dentists and dental nurses, and nuclear weapons workers) have probably been exposed to relatively low levels of mercury compounds. Few data are available on groups of workers exposed to high levels of mercury and its compounds, such as hat manufacturing workers and miners. Overall, the possibility of an increased risk of cancer of the lung, the kidney, and the central nervous system cannot be ruled out. Better epidemiologic and experimental studies are clearly needed. In particular it seems important to test metallic mercury adequately in experimental animals and to study occupational groups with substantial exposure (eg, mercury miners).