Silicone gel stimulates tumors in mice.

Silicone gel breast implants -- widely available until the Food and Drug Administration issued a marketing moratorium on them 2 years ago -- spark passionate rhetoric. Those who view them as dangerous fault FDA for not requiring rigorous testing earlier, while those who regard the implants as relatively safe chide the naysayers.

The FDA issued its moratorium on the use of the implants after reports linked them to connective-tissue and autoimmune diseases. But earlier this year, a study completed at the Mayo Clinic found little evidence to support that link (SN: 6/18/94, p.389).

Now, a new study says the same silicone gel used in implants causes a rare cancer in genetically susceptible strains of mice. The findings, reported in the July 20 Journal of the National Cancer Institute, confirm that silicone gel injected into the abdomen of certain mice will produce plasmacytomas, or tumors caused by the proliferation of plasma cells in bone and connective tissue.

"Based on the amount of data we now have, extrapolating from the mouse model to the human model is titillating but dangerous," says study coauthor Michael Potter of the National Cancer Institute in Bethesda, Md.

In humans, Potter says, plasma cell cancer appears infrequently The most common manifestation of a human plasma cell proliferation is monoclonal gammopathy of undetermined significance (MGUS), which produces benign tumors. Potter calls MGUS "an immunological wart" because it isn't usually dangerous; however, MGUS can develop into multiple myeloma, which is.

Multiple myeloma -- cancer of bone and marrow plasma cells -- accounts for 1 percent of all human cancers. It rarely spreads from bone to other parts of the body and is found most commonly in people between ages 50 and 60.

In mice, however, plasma cell cancer remains in the peritoneal cavity This distinction makes the authors wary about drawing hasty conclusions. "The tumors in the two species have many similar properties," says Potter, "but they have many dissimilar properties as well. The mouse is not a true model ... of multiple myeloma, although there are many interesting similarities between the two."

Cindy Pearson, program director of the National Women's Health Network in Washington, D.C., says the study "seems reasonably balanced" but adds that "sometimes rats and mice aren't accurate predictors. But it certainly isn't something we should ignore."

Says Sydney E. Salmon of the University of Arizona College of Medicine in Tucson, "finding a few cases doesn't mean there's a causal relationship. About 1.3 million women have received breast implants, and the incidence of myeloma is in the range of 3 per 100,000 per year. You can then calculate that over a period of 3 to 4 years, 170 women who have silicone breast implants will develop myeloma" unrelated to the implants.

Even so, in an accompanying editorial, Salmon, and Robert A. Kyle of the Mayo Clinic in Rochester, Minn., call on women who have received silicone implants and have been diagnosed with MGUS, multiple myeloma, or other monoclonal gammopathies to report their diagnosis to the FDA. If warranted, a proper epidemiological study should be conducted Salmon says. "I have no doubt some cases will be identified -- if nothing else, just on chance alone."