Over the last 30 years, scientists have worked hard to identify substances in the home, workplace, and general environment that cause cancer. This is a challenging task: in the United Sates more than 100,000 chemicals are commonly used in household cleaners, solvents, pesticides, food additives, lawn care, and other products. Every year another 1,000 or so chemicals are introduced. These are single substances that do not take into account the mixtures and various combinations of commercial and consumer products to which people in the United States are exposed every day. And many chemicals may be changed to different substances by the atmosphere, water, plants, and by incineration or combustion.

Identification of cancer-causing substances is further compounded by the fact that cancer-causing substances are sometimes created during the synthesis or combustion of other chemicals. Dioxin is an example of this kind of unwanted contaminant (see Dioxins). And besides manufactured chemicals, many natural products can also cause cancer. One example is aflatoxin (see Toxins from Fungi).

Evidence for identifying cancer-causing substances comes from three sources:

• Human studies,
• Animal studies, and
• Laboratory experiments with human cells.

Evidence from each of these sources is important in helping public health officials decide whether exposure to certain substances needs to be reduced or eliminated. The more information available, the more likely it is that they will be able to identify carcinogenic substances.

The most certain method of identifying such substances is to observe whether they have caused cancer in people. Epidemiologists design studies that follow certain populations over time to observe whether a specific agent (e.g., arsenic or benzene) or exposure (e.g., sunlight or smoking) is likely to cause cancer. Environmental causes of cancer have frequently been first noticed in the workplace. This is because workers in certain occupations have higher exposures to particular chemicals and for longer periods of time than do the general population.

The International Agency for Research on Cancer (http://www.iarc.fr), an agency of the World Health Organization, classified certain occupations as associated with cancer-causing exposures because of the increased incidence of cancers in these settings. Some of these include

• Painters,
• Furniture makers,
• Workers in the iron, steel, coal, and rubber industries, and
• Workers involved in boot and shoe manufacture or repair.

This knowledge has helped these industries and public health specialists develop processes and safety procedures designed to minimize worker exposure to cancer-causing substances. The risk is less now than in previous years.

Yet health agencies would fail in their responsibility to prevent cancer if they merely documented workplace-related cancers. They would find out about cancer risks only after many people developed symptoms of the disease, sometimes 20 to 30 years after exposure. Other epidemiology studies compare the exposure histories of people who have developed cancer with comparison groups who have not developed cancer at a particular point in time. Such studies allow researchers to look at a wide range of past exposures that may have occurred in a variety of settings, not just at those in a particular occupational setting. But these studies may miss some important links between exposures and cancer. To determine what chemicals people were exposed to many years earlier, to what degree they were exposed, and which specific ones are harmful is often difficult. But we cannot test potential cancer-causing agents with people. Observational epidemiological studies, then, are the best source of data on real world exposures and often do provide important clues.

Other testing methods involving animals and laboratory experiments are also important. They allow scientists to anticipate potential cancer-causing exposures before those exposures result in large numbers of human cancers.

Mice or rats are most commonly used to test for cancer-causing substances. Rats and mice are not only smaller, easier to handle, and more economical than larger animals; in their response to carcinogens, they are generally similar to humans. Most major forms of human cancer have been reproduced in such animals through exposure to chemical carcinogens. Because the lifetime of rodents is only 2 to 3 years, they generally provide information about the cancer-causing potential of test materials relatively quickly. Special strains of mice and rats have been developed to be particularly suitable for cancer testing. However, differences in animal and human digestive physiology complicate the relevance of diet studies in animals.

As part of an ongoing effort to reduce the use of animals in cancer testing, researchers are using human cells grown in the laboratory. Cells exposed to potential carcinogens are monitored to see whether molecular changes characteristic of cancer cells develop. Besides reducing the use of animals, these kinds of studies can be done more quickly and economically and can be useful in evaluating whether to perform the studies in rats and mice. Results from laboratory experiments also provide clues to epidemiologists about which hypotheses to test in human population studies. Human observational studies evaluating the effect of exposure to formaldehyde and methylene chloride were initiated because of data from laboratory and animal studies.

In a few cases, evidence from laboratory experiments—and knowledge of the behavior of related compounds known to be carcinogenic—is strong enough to classify a chemical as a known or probable human carcinogen. Experiments using human cells helped to classify more than 200 benzidine-based dyes as human carcinogens. Benzidine had already been classified as a known human carcinogen and scientists suspected that any dye that released benzidine inside the human body would also be a human carcinogen. When human cells grown in the laboratory were exposed to a particular dye, they were able to test whether benzidine was released. Those that did were classified as human carcinogens.

In another example, one piece of data from laboratory experiments showing that it caused DNA damage in blood cells from exposed workers led to the classification of ethylene oxide (used as a starting material in the production of other chemicals and as a disinfectant and sterilant) as a known human carcinogen.

Although such studies might reduce reliance on animals in toxicology research, testing of potential carcinogens in rodents remains an important part of cancer prevention strategies. But all scientific data available for a potential carcinogen are important, and the combination of human studies, animal studies, and laboratory experiments with human cells provides scientists with the most complete understanding of chemical risks of cancer.

Key Points

• In the United States more than 100,000 chemicals are commonly used in household cleaners, solvents, pesticides, food additives, lawn care, and other products. Every year, another 1,000 or so chemicals are introduced.
• Besides manufactured chemicals, many natural products can also cause cancer.
• The combination of human studies, animal studies, and laboratory experiments provides scientists with the most complete understanding of chemical risks of cancer.
• Human studies
  • The most certain method of identifying cancer-causing substances is to observe whether those substances have caused cancer in people.
  • Environmental causes of cancer have frequently been first identified in the workplace.
  • Health agencies would fail in their responsibility to prevent cancer if they merely documented workplace-related cancers; they would find out about cancer risks only after many people developed symptoms of the disease, sometimes as long as 20 to 30 years after the exposure.
  • Other epidemiology studies compare the exposure histories at a particular point in time of people who 1) have developed cancer, with 2) comparison groups of people who have not developed cancer. Such studies allow researchers to look at a wide range of exposures that may have occurred in a variety of settings in the past—not just at those that occurred in a particular occupational setting.
  • To determine what chemicals people were exposed to many years earlier, to what degree they were exposed, and which specific chemicals were harmful is often difficult.
• Animal studies
  • Mice or rats are most commonly used to test for cancer-causing substances because they are smaller, easier to handle, and more economical than larger animals. Also, they are generally similar to humans in their response to carcinogens.
  • Of the approximately 200 agents known to cause cancer in humans, nearly all have also been shown to cause cancer in rats or mice. But we do not know how many of the several hundred other chemicals that cause cancer in animals are also human carcinogens.
  • High doses are often used to increase the ability of tests to detect cancer-causing potential. With high dosages, any potential cancer-causing effects are more likely detected, even in small groups of rodents. This is because the cancer rate among the test animals increases correspondingly.
• Laboratory experiments with human cells
  • Researchers use human cells grown in the laboratory as part of an ongoing effort to reduce the use of animals in cancer testing.
  • Such studies can be done more quickly and economically and can be useful in evaluating whether to perform additional or follow-up studies in rats and mice.