In addition to the phenomenon of metal carcinogenesis, there is accumulating information concerning the nature and relative amounts of metal ions associated with cancer tissue in humans as well as experimental animals. Comparisons have been made between the concentration of metals present in normal and in corresponding neoplastic tissue (20, 327, 543). Most work has been done on the element zinc, its content in tumors or its change in concentration during the carcinogenesis process (2, 98, 99,240, 484). The zinc content is increased in human liver of cancer patients. The uptake of radioactive zinc by experimental mammary cancer is slightly higher than that by normal tissue. Lower than normal values of zinc have been reported in acid phosphatase associated with human prostatic cancer. But the zinc content of the prostate varies in different parts of the gland and is different from that in the cancer of the prostate. During the leukemic processes, there is a constant decrease of serum zinc levels; in the same time interval no change in whole blood occurs.

The addition of either zinc or magnesium ions inhibited activity of serum alkaline phosphatase, and this effect was greater in enzymes tested from cancerous patients than in those of the normal controls. In studies with zinc or any other ion, the exact chemical nature of the compound under investigation may be important. Zinc glycinate does penetrate into Ehrlich ascites tumor cells, whereas zinc hematoporphyrin does not.

Additional studies were made on abundance, rate of turnover, uptake and excretion of many transition elements in tumorous animals. Differences were found in the distribution of cobalt and copper as well as in the uptake of cobalt in tissues. This phenomenon may reflect disturbed hematopoiesis or the dynamic state of the proliferating cells. Thus, during the formation of proteins, nucleoproteins and enzymes with thiol groups, all of which are potential chelating agents, trace elements may be fixed simply at a higher rate. The copper content of some tumors induced by dibenzanthracene increased. Paradoxically, a high copper content in the diet of rats fed 3'-methyl-4-dimethylaminoazobenzene was reported to lower the incidence of hepatomas. This effect may be an artifact and should be reinvestigated. In leukemic patients copper levels were 50% higher in nineteen out of twenty eight patients; in one patient the values returned to normal after treatment with 6-mercaptopurine. Repeats of this investigation using penicillamine may be in order. Concentrations of copper decreased when normal epidermis was transformed into squamous cell carcinoma by cholanthrene treatment.

Nickel levels in primary cancer as well as metastasis were higher than in the comparative normal tissues. Iron values in tumors show no consistent pattern. Tumors have some influence on the storage of iron in normal mammary glands, although no higher iron content was found in the cancerous tissue itself. One reference states that there is no selective uptake of radio iron in cancer; another group reports a 300% greater uptake of ferrous citrate in tumor over normal tissue. The disappearance of an iron load from the blood of some leukemic patients was found variable relative to the type of disease. Injection of iron does not arrest the development of anemia resulting from tumors. Serum levels of iron tended to be low in humans with epithelial tumors, but higher in cases of gastric cancer. Bound iron has been liberated from tissues after treatment with hydrocarbon carcinogen.

Affinity of heavy metals for experimental tumors can be demonstrated for bismuth, especially in Ehrlich ascites tumor, and for polonium in non necrotic sections of transplanted solid tumors.

Very little work appears on the lighter elements, and especially on the alkali and alkaline earth groups. Sodium and potassium excretion values were altered in cancer surgery patients, but these patients were also treated with x ray or chemotherapy. Obviously, more work is needed in this area. Hypercalcemia quite often accompanies malignant disease, especially breast tumors. Contrariwise, in intestinal cancer the calcium level drops and the potassium level becomes elevated. Perhaps this change is related to the low adhesion of tumor cells and their increased ability to invade surrounding tissue. Calcium content of normal epidermis is decreased as it is transferred to squamous cell carcinoma by 20-methylcholanthrene.

The reports on magnesium are mainly concerned with plasma levels or the content in red blood cells; normal average values are 20.1 p.p.m. and 74.3 p.p.m. respectively. Comparative data for magnesium content in normal liver and hepatomas are available. The regenerating liver has a lower, and the hepatoma a higher, concentration of this element. Lacking are good comparative studies on the calcium magnesium ratio or calcium potassium ratio in neoplastic and normal tissue. In view of the importance of magnesium in nucleic acids, much more work should be done to compare the growing tumor with the related normal tissue. The rate of change of magnesium content during the carcinogenesis process as well as during growth of transplanted tumors may be significant. Experiments on the administration of some absorbable magnesium compound to rodents with developing tumors should be conducted.

The literature in the field of metals and tumors is mainly concerned with the metal content of the liver or perhaps the kidneys of cancer patients or experimental animals. These livers may contain higher than normal content of

Mn, Pb, Fe, Be, Cr, Ni, Si; the lungs may concentrate

Al, Pb, Cd, Fe, Cr, Si, Hg; and the kidneys may contain an excess of Cd, Mo, Cr, Hg.

These organs may have concentrated these elements prior to the disease. In experimental neoplasms, the livers of rodents with tumors had a higher copper content than those of the normal controls. This may also be true for lungs and kidneys.

Recently, melanin has been described as a cation exchange material, and the incorporation of metals in melanotic tumors was found to be parallel to the tumor incidence.

The influence of different levels of a standard salt diet on the time of development and incidence of spontaneous mammary carcinomas in mice was studied, and no significant differences were noted between treated and control animals. Addition of the ash of birch trees which contained a variety of trace metals had no effect on the growth of experimental tumors.