3866

Most women with familial breast and ovarian cancer carry germline mutations in BRCA1. Epidemiological studies strongly suggest that menstrual cycle activity is associated with increased risk of sporadic ovarian carcinoma. We hypothesized that BRCA1 mutations might predispose to gynecological cancers indirectly by influencing the menstrual cycle. To test our hypothesis, we utilized the Cre-LoxP approach to inactivate Brca1 specifically in ovarian granulosa cells, which play a critical role regulating menstrual cycle by secreting a number of reproductive hormones. We previously reported that 60% of mutant mice developed benign serous ovarian and/or uterine cystadenomas. These cystic tumors did not develop in granulosa cells and resembled human serous cystadnomas, which are benign tumors composed of the same cell type as ovarian serous carcinomas.

The estrus cycle is composed of four stages: pre-estrus, estrus, metestrus and diestrus. We sought to determine whether Brca1 inactivation in granulosa cells influences this cycle in our mouse model by measuring and comparing the length of each phase in a group of age-matched mutant versus normal mice when they were 3 months, 8 months and 14 months old. The mice were sacrificed at 14 months in order to correlate any cyclic change to tumor predisposition. Daily vaginal lavages were obtained over a period spanning several consecutive cycles. The samples were spread on glass slides and stained with Papanicolou (PAP) stain. The average length of each phase was calculated for each individual mouse and used to calculate the average length ± SD for the entire group.

An increase in the length of proestrus was seen in mutant mice that was most significant at 8 months (1.6 ± 0.45 days versus 0.9 ± 0.23 days, P=0.003). Proestrus corresponds to the follicular phase of the human menstrual cycle. There was also a decrease in the length of metestrus (corresponding to the luteal phase in the human menstrual cycle) resulting in a shorter time period between consecutive estrus phases in the mutant mice. In the 8 month old age group, mutant mice that developed cystadenomas underwent a total of 6.4 ± 0.83estrus over a 30 days period compared to 4.0 ± 0.01 in mutant mice that did not develop tumors (p = 0.03), suggesting a correlation between cycle abnormalities and tumor predisposition. Furthermore, 75% of mice that showed cycle abnormalities developed cysts compared to 33% of those who had normal cycles.

Given that proestrus is characterized by estrogen secretion while estrogen and progesterone are both secreted during metestrus, the results suggest that mutant mice have a relative increase in estrogen stimulation unopposed by progesterone. This raises the possibility that similar hormonal changes, if present in women with germline BRCA1 mutations, might account for predisposition to breast and reproductive cancers in these individuals.

98th AACR Annual Meeting-- Apr 14-18, 2007; Los Angeles, CA