A complex interplay between three different genes may determine how prostate cancer will progress, a new mouse study finds.

Genetics clearly plays a role in prostate cancer, but researchers are still working to clarify that role. Studies have shown that from 5 percent to 10 percent of prostate cancers are hereditary; the risk is doubled or tripled for a man with a brother or father who has had the malignancy, and genetic mechanisms are estimated to be responsible for half of the early-onset cases, those that occur before the age of 55.

Studies have also identified six separate genes that appear to be involved in prostate cancer. Researchers at the Robert Wood Johnson Medical School in New Jersey have been working with mice genetically engineered to have lack two of them -- NKX3, a prostate-specific gene; and PTEN, which acts to suppress the growth of tumors.

Their latest research, reported in this week's issue of the Proceedings of the National Academy of Sciences adds a third gene to the mix -- p27kip1 -- which acts to suppress tumors. Below-normal function of p27kip1 is associated with a poor outcome for human prostate cancer patients.

The results appear paradoxical. The mice with one normal p27kip1 gene showed faster progression of their prostate cancers. But cancer growth was inhibited in the mice with no p27kip1 gene activity.

"A very significant part of this report is the realization that you have to have at least some function of the p27kip1 gene to have cancer progression," said Dr. Robert D. Cardiff, a professor of pathology at the University of California at Davis, a long-distance member of the research team.

It is a puzzling finding, and "I don't envy anyone who has to explain it," Cardiff said. "It is not a simple yes-or-no answer."

Xuesong Ouyang, a part of the New Jersey team, cautions that the mouse results might not be directly applicable to human prostate cancer. "Maybe the same thing is true of human cancer, but we can't say that yet," he said.

Cardiff's view is different, and is based partly on basic genetics and partly on what he saw when he dissected the mice and studied their cancers in detail.

"This verifies that you can produce a lesion in mice using the same genes that have been documented in carcinogenesis in human prostate cancer," Cardiff said.

And, he added, "these tumors look remarkably close to what you anticipate in human prostate cancer, even though they are in a mouse prostate, which is very different from the human prostate."

The study suggests that a better understanding of how cancers progress in humans requires study of the interactions within groups of genes, rather than of single gene activity, the researchers concluded.