That certain elements in the free or combined state can be primary causative agents for human cancer has been known for many years. However, in a review of metal carcinogenesis, it is essential to consider the physical state of the agent being tested. It is also essential to accept the fact that foil implant experiments may not be valid techniques for assessing carcinogenicity.

Before classifying a metal as a true carcinogen, one must evaluate the nature of the salt used, the experimental procedure, the species tested, and the amount of material needed to cause cancer. If an agent causes tumors in only one species only under special conditions, should it be classified as a carcinogen? For example, there may be a good possibility that what is reported as metal carcinogenesis could be the induction of tumor by irritative or traumatic means, such as the carcinogenic effect of ferric oxide. If the cancer forms as a result of a large excess of a normal metallic ion upsetting ionic equilibria, one can question whether this is true carcinogenesis. Cancer of the lung among asbestos workers may also be the result of an irritative agent. Excluded from this discussion are those radioactive materials, such as some uranium compounds, which presumably induce tumor by an irradiation mechanism. The early work in this field has been summarized by Stem and Willheim.

A summary of the elements reported to be carcinogenic appears in a table in the appendix. Each row of the periodic chart is represented. Beryllium (At. No. 4) is the first element encountered in the periodic chart which is carcinogenic; osteosarcomas have been found in rabbits after administration of beryllium compounds by the intravenous route. Workers exposed to these compounds have dermatological and respiratory tract ailments. Although no human tumors have been reported after exposure to this element, these workers should be followed to see if lung carcinoma develops in the future. A reference has been found that lists aluminum (At. No. 13), a second row element, as also having carcinogenic properties. There are no recent experimental reports substantiating or refuting this claim; this should be checked.

As to be expected, the majority of the well established carcinogenic metals are found in the transition elements. Chromium (At. No. 24) has been implicated many times. An occupational hazard of chromate workers is lung cancer, and tumors can be induced in rodents experimentally by chromium compounds. No references could be found on carcinogenic tests for vanadium (At. No. 25). Iron (At. No. 26) as iron dextran produced sarcomas at site of injection in rats after weekly administration of the compound. These observations were confirmed in mice and rats, whereas uncom plexed dextran controls were negative. Sarcomas are now being reported as appearing in humans at the site of injection. In a sense this was predictable.

Recently cobalt metal (At. No. 27) has been found quite toxic for rats, and even carcinogenic. The oxide of cobalt obtained from the flue of a nickel refinery produced rat sarcomas which metastasized. The control element, tungsten (At. No. 74), did not produce tumors.

There is a greater incidence of lung tumors among nickel (At. No. 28) mine workers than among the population as a whole. The same is true for those working with nickel alloys, and nickel sulfide recovered from the refinery flue was suggested as the carcinogen. In rats sarcomas can be produced by implanting nickel powder, or by injecting an organometallic compound. EDTA has been reported as having some protective action against nickel carcinogenesis, presumably by liberating bound histamine. Is it possible that EDTA acts by sequestering and thus removing nickel from action?

Nickel carbonyl [Ni(CO)4] is one of the most toxic gases encountered in industry. After some inconclusive studies of the possible carcinogenic activity of this gas in rats, a species very resistant to the induction of lung tumors, it was found that four out of nine rats did develop pulmonary tumors after prolonged exposure to this agent. This experiment renews interest in the possibility that nickel may be a carcinogenic agent in tobacco smoke. No study to date has definitely proved that smoking causes cancer in humans or animals. Circumstantial evidence would seem to implicate nickel. Cigarettes contain an average of 2 fig. of nickel per cigarette; there is in the main stream smoke 2-7% of carbon monoxide. If nickel and carbon monoxide combine to form nickel carbonyl, then in one year a man will inhale three times the amount of Ni(CO)4 reported to be carcinogenic to the rats. Chemically there are objections to this. In the laboratory Ni(CO)4 can only be made from metallic nickel and carbon monoxide and once formed, decomposes at temperatures just above 100° C.

Studies should be made on the nickel content of the cigarette ash to see what amount of this element transfers to the mainstream smoke. The carcinogenicity or cocarcinogenicity of other elements should be studied. The high accumulation of aluminum in lungs may be a factor to consider.

Most of the experiments involving copper (At. No. 29) deal with salts or chelate compounds that were tested for anti tumor properties. Yet copper sulfate when injected produced tetratoma of the testes in the rooster during high gonadal activity. In the same column as copper is silver (At. No. 47), and it also has been reported carcinogenic, causing sarcomas at site of implantation of foil. A compound of zinc (At. No. 30), zinc chloride, produced the same type of tumor in fowls and rats as did copper sulfate. Metallic mercury (At. No. 80) in the same family was productive of tumors. Cadmium (At. No. 48), listed between zinc and mercury, was conspicuous by its absence. It is quite toxic and inhibits many enzymes. Just before this manuscript was completed, the carcinogenicity of this element was reported. Given intramuscularly in mice or rats, a severe inflammation occurred three days after injection, and rhabdomyosarcomas appeared 23-24 weeks later. Although germanium is not mentioned in the voluminous literature on carcinogenesis, tin (At. No. 50), a member of the same column, did produce tumors in rats. In that same experiment lead (At. No. 82) was also tested but not found active. A recent report disclosed the finding of a cerebral tumor in a lead worker. This may be a coincidence, or may be the forerunner of other reports.

Among the first reported exogenous carcinogens were arsenic (At. No. 33) and its compounds. Most of the earlier work on arsenic and cancer is available in a review. A retrospective study in epidemiology of arsenic poisoning and cancer has just appeared. The role of arsenic must also be evaluated in a consideration of the possible relation of smoking and human lung cancer. In the years 1939-1956, the arsenic content of cigarettes manufactured in Great Britain was between 7-51 p.p.m. calculated as arsenic oxide, As203. Since that time the residue values have fallen below those of 1939. The last element in the third row of elements found to cause cancer is selenium (At. No. 34). Hepatomas were found in rats following ingestion of more than minute amounts of selenium. Cirrhosis preceded the appearance of tumors.

It is important to realize that the number of inorganic agents found to be carcinogenic is small compared to the many tested. With few exceptions each of these carcinogens also has some specific biological function. The mechanism of action may involve the antagonism of one essential ion by another. In large excess, the carcinogenic ion may equilibrate with the natural metal coenzyme and inactivate or retard the cata bolic reactions; or more likely, these metal combinations may enhance the kinetics of normally acting enzymes. Examples of the latter are well known. It is doubtful if the abnormal metal functions solely by displacing the essential metal in a metal dependent enzyme system.