Diabetes drug stirs cancer confusion.

New research has offered hope that troglitazone, a widely prescribed diabetes drug, could become a new colon cancer therapy, but some scientists also express concern that the drug may trigger the cancer in genetically susceptible people.

This seemingly contradictory story emerges from several studies in the September NATURE MEDICINE. In one, Bruce M. Spiegelman of the Dana-Farber Cancer Institute in Boston and his colleagues report that troglitazone and other compounds that activate an intracellular protein called PPAR-gamma can transform colon cancer cells into nonmalignant cells in test-tube experiments. The drug also inhibited growth of tumors formed from human colon cancer cells implanted into mice.

PPAR-gamma normally regulates cell proliferation and specialization. Spiegelman and his colleagues hope soon to test whether troglitazone can help patients with colon cancer. The drug has already shown some promise in people with liposarcoma, a different cancer, says Spiegelman.

Curiously, however, two research groups have found that PPAR-gamma activators increase the incidence of cancer in certain mutant strains of mice. One group, led by Ronald M. Evans, a Howard Hughes Medical Institute investigator at the Salk Institute for Biological Studies in La Jolla, Calif., tested such compounds on mice with mutations in a tumor-suppressor gene called APC. These animals are generally predisposed to cancer of the small intestine. People with APC mutations almost invariably develop cancer of the colon.

The scientists had hoped that the PPAR-gamma activators might reduce cancer risk, but treated mice developed many more precancerous polyps in the colon than untreated ones. "What we found was exactly opposite what we had expected," says Evans. Yet normal mice receiving the PPAR-gamma activators had no greater incidence of cancer. Similar results have emerged from studies conducted on APC mutant mice by a group led by Johan Auwerx of the Pasteur Institute in Lille, France.

Since fats can activate PPAR-gamma, these studies may explain the finding that diets high in fat increase a person's risk of colon cancer. Evans also warns that people with diabetes who take troglitazone; especially those with a family history of colon cancer, should be carefully monitored for any signs of the disease. Spiegelman, however, notes that the mice may not reflect how colon cancer develops in people and that only those with APC mutations may face an increased cancer risk. "There's no evidence that [troglitazone] induces polyps in normal people or mice," he says.