The scientists believe that constant ovulation, which causes cells in the ovary to divide, is likely to spontaneously damage DNA in those cells over time. That can result in mutations to a critical regulatory gene, known as p53, that normally stops cells from proliferating into cancer.
"When a woman ovulates, the egg that is extruded from the ovary blows a hole in the surface of the ovary," said co-author Dr. Andrew Berchuck, professor of gynecologic oncology. "The epithelial cells, where the cancer starts, line the surface of the ovary and they have to proliferate to fill in the hole."
The findings indicate that women at risk for this type of cancer can protect themselves by reducing their ovulation cycles through birth control pills, pregnancy and breast-feeding, the Duke researchers said in the study report, published in the July issue of the Journal of the National Cancer Institute.
The research was funded by the Centers for Disease Control, the National Cancer Institute, and the American Cancer Society.
Ovarian cancer is usually a fatal disease because it is usually found in its late stages, after it has spread beyond the ovaries. In 1996, about 27,000 new cases were diagnosed, and about 14,800 women died of the disease, according to the National Cancer Institute. Mean survival after detection is less than 3 years.
Tumors that show evidence of p53 genetic mutation account for half of all ovarian cancer cases, and are considered the most aggressive form of ovarian cancer.
The study helps clear up the mystery of why it seems that some women, but not all, can protect themselves against ovarian cancer by suppressing ovulation, the investigators say.
"We found that there is a distinct subgroup of ovarian cancer, those associated with p53 mutations, that is linked to ovulation," said lead author, Joellen Schildkraut. "This strongly suggests that women at risk for this type of cancer can protect themselves by suppressing their ovulation."
The study also means that the rest of ovarian cancers are likely caused by "different pathways" that aren't linked to ovulation, said Schildkraut, assistant professor in community and family medicine and a researcher at Duke Comprehensive Cancer Center. Some of those pathways are due to different genetic mutations.
"This is one of the first studies to show that ovarian cancer is really a collection of a number of different cancers," said Berchuck. "Each subgroup will have its own risk factors, which can help physicians tailor protective strategies and specific treatments."
The Duke study is novel because it takes a "molecular epidemiology"approach to the disease, in order to link risk factors seen over a population of patients to alterations in the biology of the cell, Berchuck said. "We are trying to understand on a cellular basis why cancer behaves differently in different patients."
For this study, the researchers used one of ovarian cancer's "molecular signatures" -- over-expression of the p53 gene in cell tissue. Such over-expression is characterized by abnormal accumulation of the gene's protein and occurs when that gene is mutated.
They then tested for the presence of p53 mutations in samples of ovarian tumors taken from 197 cancer patients. These women had participated in the Cancer and Steroid Hormone (CASH) study, which also included a "control" group of 3,363 cancer-free women. All of the women were interviewed to collect data on their socioeconomic status, age at menarche (beginning of ovulation) and menopause, use of oral contraceptives, and other hormones, infertility, pregnancy, breast-feeding, medical history, family history of cancer, as well as other information. For each woman, the number of lifetime ovulatory cycles (LOC) was computed.
The formula derived was based on an average of 13 cycles per year (each 28 days long), minus the months a woman was pregnant, breast-feeding or using birth control pills. Variations in a woman's cycle -- whether she ovulated more or less than the typical 13 times yearly -- was not considered. An average woman starts ovulating at age 12.5 and enters menopause at 50, so her unsuppressed number of lifetime ovulations is about 487.
Schildkraut, Berchuck and co-investigator Eugenia Bastos, also from Duke, then compared the LOC of women with cancer that over expressed p53 protein (p53-positive) to women whose cancer did not have a p53 signature (p53-negative). Estimating that 235-375 cycles is a "medium" LOC and 376 to 533 is a "high" LOC, they found:
Women with p53-positive tumors were seven times more likely to have had a moderate number of LOCs and 7.7 times more likely to have had a high number of LOCs than women with p53-negative tumors.
When data were controlled for age, menopausal and child-bearing status, women with p53-positive tumors were 4.3 times more likely to have had a moderate number of LOCs and 9.1 times more likely to have had a high number of LOCs compared with the cancer-free control group.
"This supports what other epidemiological studies have postulated -- that suppression of ovulation due to breast-feeding, number of pregnancies, and use of oral contraceptives decreases the risk of this type of cancer," Schildkraut said in an interview. Berchuck added that the results show "a dose response. The more you do of each increases your protection."
Additionally, women who have longer cycles than the typical 28 days have less exposure to DNA damage, while women whose cycles are shorter are at a higher risk because they ovulate more over a lifetime, Berchuck said.
Berchuck said the study provides the "first real evidence of how birth control pills work at a cellular level to prevent ovarian cancer."
The researchers noted that the 197 patients included in the study ranged from 20 to 54 years of age, but that the mean age of onset of ovarian cancer is 60. They said further study of the association between ovulation and p53-associated cancer is needed in a large group of older women. "If our data show this kind of association in younger women who haven't yet completed their ovulatory lives, it makes sense that total LOCs will be even more powerful a risk factor," said Berchuck.
The study also suggests that development of a drug that specifically promotes cell death (apoptosis) in ovarian tissue after ovulation might help women reduce their risk of DNA damage, but still allow them to ovulate. "There is some evidence that causing these cells to die protects against ovarian cancer because it eliminates mutated cells," Schildkraut said.
Berchuck's presentation of the study's preliminary findings at the Society of Gynecologic Oncologists' annual meeting in March was cited as the best oral presentation and he was given the society's "President's Award."