Shape-altering genes linked to ovarian cancer

Researchers have explained how and why different forms of ovarian cancer evolve in a discovery that could lead to earlier detection and perhaps more personalised treatment for a disease that will claim an estimated 16 000 women's lives in the US in 2005.

Frequently referred to as a silent killer, ovarian cancer offers few clues to its presence, often until it has spread beyond the ovary to other tissues. Early detection has been difficult because ovarian cancer is not a single disease, but appears in many forms, with each form behaving differently.

The American Cancer Society estimates that about 22,220 new cases of ovarian cancer will be diagnosed in the United States during 2005. A woman's risk of getting ovarian cancer during her lifetime is about 1 in 58, accounting for about 3 per cent of all cancers in women.

Researchers have discovered that a set of shape-altering genes become activated in ovarian cancer. These HOX genes, better known for their role in normal embryonic development, direct the cancer cells to take a variety of different forms, depending on which of the genes is turned on. The researchers reported their finding in the April 10, 2005 on-line issue of the journal Nature Medicine.

Honami Naora, an assistant professor in the University of Texas' Department of Molecular Therapeutics said: "Our finding explains how each of the three major forms of ovarian cancer acquire their unique appearance."

"These genes cause a metamorphosis of the ovarian epithelial cells, directing them to change their shape," she said

These shapes make each ovarian cancer different from each other, and also different to the surface epithelium or outer covering of the ovary from which these cancers are thought to arise. Serious ovarian cancer exhibits features resembling those of the fallopian tubes, the endometroioid form has features resembling the lining of the uterus; mucinous ovarian cancer even looks like intestinal cells.

These mysterious shapes have caused some researchers to speculate that ovarian cancers might originate from some other tissues, and not the ovarian surface epithelium at all. Naora reasoned that ovarian tissue might be coaxed into the different forms by changes in its genetic programming.

The HOX genes, which direct immature embryonic tissue to form the various body structures during development, could become reactivated in ovarian cancer cells. Naora's team tested the effect of four HOX genes on cells derived from the ovarian surface epithelium and found that activating different HOX genes caused the cells to change into different shapes and resemble the forms seen in ovarian cancer.

For example, turning on HOXA9 caused cells to form tumours that resembled high-grade serous ovarian cancer. On the other hand, HOXA10 activation resulted in tumours that resembled endometrioid ovarian cancer and HOXA11 caused cells to form tumours that resembled mucinous ovarian cancer.

In addition, the research team found that activation of HOXA7 in combination with any of the other HOX genes resulted in formation of low-grade tumours that are less aggressive than high-grade tumours.

Because HOX genes are sensitive to levels of the female hormones oestrogen and progesterone produced in the reproductive organs, Naora speculates that abnormal changes in levels of these hormones could explain how the HOX genes come to be turned on in ovarian tissue. Indeed, most of the known risk factors for ovarian cancer are related to levels of these same hormones.

"One of the major problems with diagnosis and treatment of ovarian cancer is that it is not a single disease," Naora says. "Each form of ovarian cancer has its own unique clinical behavior. If we understand what causes these different forms, we have taken the first step toward early diagnosis and therapy tailored to each of the various subtypes."

The eventual goal of Naora's research is to develop a molecular profile or pattern that could be used to determine who is at greatest risk of ovarian cancer and to tailor treatment for women who do develop ovarian cancer.

"An impediment to improving early detection of ovarian cancer is the lack of well-defined pre-malignant or precursor lesions. Furthermore, right now we don't have any way to assess the relative risk for women with a strong family history of ovarian cancer," she said.

"Often these women have their ovaries removed, and unusual changes in the cell shape are seen in many cases. But we don't know if changes in cell shape necessarily lead to cancer. We would like to be able to offer a test that could assess risk and allow women to make more informed choices," she added.