Dr. Ray DuBois Jr.
VUMC group finds new way to combatovarian cancer
S.K. Dey, Ph.D.
A surprising new scientific finding by researchers at Vanderbilt University Medical Center may hold answers to better treatment of ovarian cancer and will certainly alter the course of cancer research.
Their study, published last Saturday in the journal Cancer Research, shows that the enzyme COX-1, and not the COX-2 enzyme, which is the current target of therapy, is expressed in tissue samples from women with epithelial ovarian cancers.
This finding suggests a new direction in cancer care — using drugs that aim to block both COX-1 and COX-2 or which block COX-1 alone. "This could have a significant clinical impact with regard to our thinking about ovarian cancer," said Dr. Raymond N. DuBois, Mina Cobb Wallace Professor of Gastroenterology and Cancer Prevention.
Ovarian cancer is the most lethal gynecological malignancy worldwide. The American Cancer Society estimates 25,400 new cases in the U.S. in 2003. More than half of these women will die of the disease because it has few symptoms and is usually not detected until it is well advanced.
The Vanderbilt researchers studied 11 tissue samples from women diagnosed with ovarian cancer and nine normal ovarian tissue samples. The cancerous tissues had not been exposed to chemotherapy, which is known to induce COX-2 expression.
"Significant levels of COX-1 or COX-2 were not detected in any of the normal human ovarian tissue," the scientists wrote. "However, dramatic elevations of COX-1, not COX-2, protein and mRNA were detected in a majority of the ovarian cancer samples tested."
“I couldn’t believe it. We were shocked,” said co-investigator S.K. Dey, Ph.D., Dorothy Overall Wells Professor of Pediatrics and professor of Cell & Developmental Biology and Pharmacology.
This finding was contrary to previous reports, which found COX-2 expression in ovarian cancer. Those studies, however, did not exclude tissue samples that had been previously exposed to chemotherapy, DuBois said.
The Vanderbilt researchers also found high levels of the COX-1 enzyme in human ovarian cancer cells grown in culture. “When we added a COX-1 selective inhibitor, we could block the ability of those cells to make angiogenic growth factors,” DuBois said.
Angiogenic growth factors stimulate in neighboring tissue the growth of blood vessels, which nourish cancer cells and aid their rapid growth.
DuBois cautioned that their work was preliminary and that further study is needed. However, there is serious discussion nationally about testing COX-2 inhibitors as a potential treatment for ovarian cancer. “With this finding, I’m wondering if it’s the right thing to do … until we have more information,” he said.
In addition, some women who have had ovarian cancer may be taking selective COX-2 inhibitors in an attempt to reduce the risk of a recurrence of the disease. “We can’t make any recommendations on that because there’s no FDA approval, but it may be that if they’re going to do that they should take a non-selective instead of a selective inhibitor because then you hit … both (enzymes),” he added.
The cyclooxygenase (COX) system catalyzes the production of powerful substances called prostaglandins. These lipid mediators exert a wide range of effects, from inflammation and healing wounds, to forming blood clots and promoting cancer growth.
Non-selective inhibitors, which block both COX-1 and COX-2, include non-steroidal anti-inflammatory drugs like aspirin and ibuprofen. Celebrex and Vioxx are selective COX-2 inhibitors approved for treatment of arthritis and pre-cancerous colorectal cancers in certain patients.
DuBois, who also is associate director for Cancer Prevention in the Vanderbilt-Ingram Cancer Center and a member of the National Cancer Institute’s Board of Scientific Advisors, helped establish the relationship between COX-2 and colon cancer. Selective COX-2 inhibitors are now being tested as potential treatments for a wide variety of cancers.
Dey came to Vanderbilt last summer from the University of Kansas Medical Center to head up the new division of Reproductive and Developmental Biology within the Department of Pediatrics. His research has helped advance the understanding of female fertility regulation.
Dey and DuBois started collaborating in the late 1990s, when both were investigating role of COX-2 enzyme and angiogenesis. While DuBois was searching for clues to cancer growth, Dey, a reproductive and developmental biologist, was trying to understand how the enzyme is involved in implantation of the embryo in the uterine wall – the pivotal event for a successful pregnancy.
Just as tumor cells stimulate the growth of blood vessels in surrounding tissue, the embryo can stimulate blood vessel growth in the uterine wall. “There is a lot of similarities,” Dey said. “Not only (is) COX-2 … highly expressed at the site of implantation, we have shown that COX-2 is involved in angiogenesis.”
Dey had been using genetically engineered mice that lacked the COX-2 enzyme, and he and DuBois collaborated on a study that helped establish the role of the enzyme in angiogenesis. To continue their current studies of ovarian cancer, Dey and his colleagues are developing mice that over-express either COX-1 or COX-2 in their ovaries, but not both enzymes.
The goal is to learn why ovarian cancer, but not other forms of the disease, may express high levels of COX-1. One possibility: reproductive hormones secreted by the pituitary gland may influence enzyme expression, Dey said.
Rajnish A. Gupta, Ph.D., a student in the Vanderbilt Medical Scientist Training Program in DuBois’ lab, was the first author on the Cancer Research paper. Other co-authors included research assistant Lovella V. Tejada in Dey’s lab; Kansas City, Mo., gynecologist Dr. Beverly J. Tong; Sanjoy K. Das, Ph.D., associate professor of Pediatrics; and Dr. Jason D. Morrow, F. Tremaine Billings Professor of Medicine and professor of Pharmacology.
The study was supported by the National Institutes of Health and the National Colorectal Cancer Research Alliance.