While panel members said these steroids have important medical uses and clear medical benefits, they have long been associated with a risk of uterine, endometrial and breast cancers. The panel agreed 8 to 1 that these hormones cause an elevated risk and should be considered not merely as associated with increased cancers but as substances that are "known to be a cause of human cancers."

The federal Report on Carcinogens is required by Congress to inform the public, medical community and regulatory agencies about potential cancer-causing substances. It is prepared by the National Toxicology Program, which is headquartered at the National Institute of Environmental Health Sciences in Research Triangle Park, NC. NIEHS/NTP sought the views of the panel of 13 outside scientists as one step in the development of the Tenth Report on Carcinogens, which will be written and published after further public comment and review.

Estrogens occur naturally in women and to a lesser degree also in men. They have important medical uses as estrogen replacement therapies in post-menopausal women and for birth control. Their use has long been associated with a risk of uterine, endometrial and breast cancers. They are so labeled, panelists said, and doctors and women should weigh their known benefits against these risks. There was no suggestion by the panelists that estrogen use be restricted or eliminated.

Another medical product, the antibiotic chloramphenicol, was recommended for listing as "reasonably anticipated" to be a cause of human cancer. Found to be effective against typhus in 1948, it was one of the first antibiotics in large-scale production but after being linked to a risk of potentially fatal aplastic anemia, the drug's use was generally limited to cases of typhoid fever and meningitis when other antibiotics were ineffective.

The scientific panel, a part of the National Toxicology Program Board of Scientific Counselors, made its recommendations after three days of discussions (Dec. 13-15) in Washington, D.C.

It also recommended for listing as "known" human carcinogens:

• Common wood dust produced in sanding furniture and cabinets, which is associated in industrial settings with increases in cancer of the nasal cavities and sinuses. About two million people world-wide are exposed to wood dust routinely in their jobs.

• Broad spectrum ultraviolet (UV) radiation from sunlight or artificial sources. However, the individual classes UVA, UVB and UVC were recommended for listing in the different category of "reasonably anticipated to be human carcinogens" because of difficulties in separating overlaps between classes.

Recommended for listing as "reasonably anticipated" to be a human carcinogen (a second category, used generally when strong, causal human data or mechanisms have not been shown):

• Methyleugenol, a naturally occurring flavor in basil, cinnamon leaves, nutmeg, mace, pimento, bananas, black pepper, bilberries and blackberry essence. It is also added as a flavoring agent in minute quantities to some jellies, baked goods, nonalcoholic beverages, chewing gum, candy, pudding, relish and ice cream. Methyeugenol has been linked to liver, glandular, kidney and mammary gland tumors in laboratory animals but there have been no studies as yet to show if it causes cancer in humans, NIEHS/NTP staff said.

• Metallic nickel as encountered in some industrial uses (recommended by a split 7-3 vote but the panel voted 9 to 1 against also listing certain alloys of nickel as well. The alloys have many uses as specialty steels and stainless steels and are used in some medical implants.

Trichloroethylene (TCE), a degreaser for metal parts such as those used in manufacturing metal products, electrical and electronic equipment and for cleaning transport equipment, was considered for upgrading to "known" human carcinogen from "reasonably anticipated" but the panel recommended that the chemical continue in the second category.

The panel recommended, 7 to 3, that ordinary talc not be listed as either known to be or reasonably anticipated to be a human carcinogen. The panel reviewed a series of studies of women with ovarian cancer but was not convinced that the excess cancers could clearly be related to genital talc use.

A particular type of talc, called talc with asbestiform fibers, contains fibers that have an appearance similar to asbestos but are not truly asbestos. The panel rejected a recommendation that these be listed as "known" human carcinogens, and the panel split, 5 to 5, over whether the data on lung cancers in talc miners and millers justified a listing as "reasonably anticipated to be a human carcinogen."