were inadequate. Subcutaneous injections totaling about 6 mg/kg of sulfur mustard produced sarcomas and other malignancies at injection sites in C3H, C3Hf, and strain A mice, but did not produce an increase of malignancies at other sites.

Nitrogen mustard, particularly HN2, was more widely tested than sulfur mustard and was found to be a potent tumorigen and carcinogen. It produced an increase in lung tumor nodules in strain A mice from both intravenous and intraperitoneal injections. Subcutaneous exposures produced injection site tumors and pulmonary tumors in selected strains of mice. Its tumorigenic potency appeared to be similar to or somewhat greater than sulfur mustard; HN2 was one of the most potent alkylating agents tested in the strain A bioassay program. HN2 increases the action of both X and UV radiation. Excess lymphosarcomas were reported in two studies of this type, and thymic lymphomas in a third. However, the numbers of animals used in these studies were small.

Epidemiologic evidence about the carcinogenicity of mustard agent and Lewisite comes from three sources: (1) studies of soldiers who were exposed to these agents on the battlefield; (2) studies of workers who manufactured these agents; and (3) studies of the unintended adverse effects on patients exposed to these agents for medical therapy. This review describes key studies from each of these arenas and characterizes their attributes, including study design, comparison population, case ascertainment and definition, control of other (confounding) important variables, and estimation of exposure to mustard and Lewisite.

Japanese Studies. A number of studies have followed workers at a weapons factory on Okuno-jima, an island in the Inland Sea and in the south of Hiroshima Prefecture. The factory manufactured various chemical agents for use in World War II, and many investigators believe that these workers were exposed to multiple poisonous agents other than sulfur mustard and Lewisite, including HCN, diphenyl cyanarsine, choroacetopheonone, and phosgene. The studies reflect use of various methodologies in epidemiology ranging from case reports and case series to formal cohort mortality studies. The weight of the evidence firmly demonstrates the carcinogenicity of sulfur mustard in the occupational setting, which implies repeated exposures but does not allow comment on the level of exposure.

Yamada (1963) reported a case of bronchogenic carcinoma in a former worker in the Hiroshima plant. Other cases are included in this report,