Substantial progress toward the control of cancer has stemmed from the realization that chemical agents can cause malignancy. The first recognition that a chemical exposure in the workplace could cause human cancer was the report by Percival Pott in 1775 of an excess incidence of squamous cell carcinoma of the scrotum among young chimney sweeps in London.1 Pott attributed those tumors to soot. Later studies confirmed Pott's observation and demonstrated that the chemical agents in soot responsible for scrotal cancer in the “climbing boys” were polycyclic aromatic hydrocarbons, particularly benzo-(a)-pyrene.2 In another pioneering observation, Rehn3 in 1895 noted a high frequency of cancer of the urinary bladder among workers in the aniline dye industry. He attributed those tumors to aromatic amines. More recently, etiologic associations have been noted in working populations between benzene and leukemia4; asbestos and lung cancer5; arsenic and cancers of the skin, lung, and liver6; bis-chloromethyl ether and lung cancer7; and vinyl chloride monomer and angiosarcoma of the liver.8
Occupational exposures to carcinogens are widespread. The National Institute for Occupational Safety and Health calculates that over seven million American workers have the potential for regular occupational exposure to proven human carcinogens. However, the full extent of occupational exposure to carcinogens cannot be known at present, because only about 30 percent of the 70,000 chemical compounds registered with the Environmental Protection Agency have ever been tested for carcinogenicity.9 It is entirely likely therefore that millions of American workers are regularly exposed to chemical compounds whose malignant potential has not been defined.
Estimates of the fraction of cancer deaths that are caused by occupational exposures have varied from less than four percent,10 to 12 percent, to 20 percent11 or more.12,13 This wide range reflects lack of data on the carcinogenic potential of most industrial chemicals9 and the absence of effective public health surveillance systems for occupational disease.14 The commonly cited figure that only four percent of cancer deaths in the United States are work related is almost certainly too low. That estimate failed to consider cancers arising in persons above age 65, although cancers in that age group account for more than two thirds of all human malignancies and frequently include tumors of occupational origin.15,16
A reasonable lower-bound estimate of the fraction of American cancer deaths that are caused by occupational exposures appears to be 10 percent.17 This estimate implies that at least 10 percent of the more than 500,000 cancer deaths that occur each year in the United States are caused by toxic occupational exposures, acting either alone or in synergy with other carcinogenic exposures, most notably cigarette smoking. Cigarette smoking remains the single most important chemical carcinogen and accounts alone or in synergy with occupational factors for as much as 40 percent of all cancer in the United States. In certain subgroups of the American population, the mortality rate for occupational cancer is much greater than the national average.18 For example, more than 50 percent of asbestos workers die of asbestos-related cancer,19 and in some cohorts of chemical workers exposed to aromatic amines more than 80 percent have died of bladder cancer.20 By the year 2030, the asbestos epidemic will have claimed an estimated 300,000 lives in the United States alone.18
Two points are of great importance in occupational carcinogenesis: (1) virtually all occupational cancer is preventable,13 and (2) most occupational carcinogens have first been recognized by clinicians.21
Virtually all occupational cancers arise as a direct consequence of human activity. In theory, therefore, they can be prevented by altering this activity. Further, occupational cancer can be prevented much more easily than cancer caused by adverse lifestyle factors, such as cigarette smoking or excessive consumption of alcohol, because occupational cancer is typically the result of a point-source of common exposure — the workplace. Control of common sources of toxic exposure has successfully been achieved many times in American society.18 By contrast, attempts to alter lifestyles must seek to modify the behaviors of millions of individuals, many of whom are addicted or habituated to chemical toxins.
A most effective approach to the prevention of occupational cancer is premarket testing of all new chemical compounds and industrial processes. Unfortunately, premarket testing has often not been undertaken.9 A second approach is to take sensible measures to control toxic exposures in the workplace. Carcinogens must be replaced with safe, noncarcinogenic substitutes; hazardous processes must be enclosed and ventilated; and workers must be provided with education and counseling to enable them to protect themselves. Finally, when all else fails, personal protective equipment such as respirators and special clothing may be required.
Regulatory standards issued by the Occupational Safety and Health Administration and the Environmental Protection Agency are extraordinarily important in the prevention of occupational cancer. These standards regulate permissible uses of carcinogenic chemicals and establish levels above which workers may not legally be exposed. All standards are, however, inherently arbitrary — they imply safety when safety does not exist. There is no bright line between the level of exposure to a toxic substance that causes cancer and that which is safe; there is instead a continuum of toxicity. Standards therefore need to be continually re-examined in the light of new data and revised when necessary.
Virtually all occupational carcinogens have first been recognized by astute clinicians. The most important instrument for the proper diagnosis of occupational cancer is a carefully obtained history of occupational exposure.22,23 The occupational history is of central importance here because most occupational cancers are not clinically or pathologically distinct from cancers caused by other factors. For example, lung cancer caused by asbestos is indistinguishable from lung cancer caused by cigarette smoking. Only rare occupational cancers, such as mesothelioma caused by asbestos5 or hepatic angiosarcoma caused by vinyl chloride monomer,8 are highly unique and virtually pathognomonic. Diagnosis of occupational cancer is further complicated by the long latency period that frequently must elapse between first exposure to a toxic substance and the appearance of disease.
Physicians are in a unique position to identify cancers caused by well-known occupational hazards such as asbestos and benzene. Following such recognition they can be very effective in triggering preventive activities by industry, unions, and public authorities. In addition, physicians are in a unique position to identify new associations between workplace exposures and human malignancy. Perhaps in no field outside of occupational medicine does there exist so great an opportunity for a physician to achieve immortality through eponym.24