Pathogenesis of serous, extra-uterine Müllerian epithelial cancer and therapeutic implications.

The model illustrated in Figure 1 is based on published data and arguments reviewed in this chapter. It is well established that cancer, in general, arises in a background of cell proliferation. It is probable that the main proliferation signal for serous extra-uterine Müllerian carcinoma precursors comes from the hormonal changes associated with the menstrual cycle that act cell non-autonomously as predicted by the aforementioned cell non-autonomous hypothesis. This cell non-autonomous scenario is magnified either by increased ovulatory activity (incessant ovulation) in individuals affected by sporadic cancers, or by the consequences of decreased BRCA1 or BRCA2 gene dosage in BRCA1/2 mutation carriers. It is also well-established that extra-uterine Müllerian serous tumors almost always carry P53 mutations and that such mutations are present not only after acquisition of invasive and metastatic ability, but also in precursor lesions ( 85 ). The presence of such mutations in cells subjected to growth stimulation (via the cell non-autonomous pathway driven by the ovulatory cycle) inevitably leads to accumulation of chromosomal abnormalities due to absence of important cell cycle checkpoints controlled by P53, as observed in the in vitro model described by Luo et al . ( 76 ). Eventually, chromosomal abnormalities become overwhelming, resulting in a cell cycle arrest at the spindle assembly checkpoint ( 77 , 78 ). As illustrated in Figure 1 , two different mechanisms can lead to recovery from this mitotic arrest, both leading to aneuploidy from a polyploid intermediate. In the absence of reduced or absent BRCA1 function either due to a germline mutation or promoter methylation, cells overcome this checkpoint and return to interphase without completing cytokinesis as described by Yu et al . ( 78 ). The resulting bi-nucleated cells may eventually show a single, enlarged nucleus following subsequent mitoses as depicted in Figure 1 . In the presence of normal BRCA1/2 function, cells arrested at the spindle assembly checkpoint might overcome this checkpoint via an alternate mechanism entailing CCNE1 amplification, as suggested by the Cancer Genome Atlas dataset, which shows frequent amplification of this cyclin in the absence of BRCA1/2 mutations (see above). The proposed mechanism is that centrosome duplication, one of the consequences of CCNE1 over-expression ( 86 ), leads to increased microtubule formation and anchoring ( 87 ), resulting in disruption of the spindle assembly checkpoint. Thus, the combination of cell non-autonomous factors driven by hormonal factors or BRCA mutations, and of cell autonomous factors entailing recovery from a mitotic arrest at the spindle assembly checkpoint, leads to polyploid cell populations that invariably become aneuploid due to instability of the polyploid state. The combination of chromosomal imbalances, superimposed on accumulation of additional genetic defects provides an excellent soil for malignant transformation. The cell non-autonomous effects influence proliferation of serous extra-uterine Müllerian epithelium and account for the site specificity of the tumors associated with the BRCA mutation carrier state while the cell autonomous effects lead to the characteristic near polyploid state of these cancers. Relevance to human tumors is illustrated in Figure 2 , which shows examples of either binucleated cells ( Figure 2A ) or multipolar mitoses ( Figure 2B ) in histological photographs of representative high-grade serous extra-uterine Müllerian carcinomas.

Cell culture models mimicking the genetic background of cancer precursor lesions, such as the model developed by Luo et al . ( 76 ), can provide clues about the cellular mechanisms leading to the near polyploid state that typically characterizes serous high-grade extra-uterine Müllerian carcinomas. These authors used cultures of cystadenomas, the benign counterparts of extra-uterine Müllerian carcinomas, in which SV40 large T antigen was forced-expressed. These genetic manipulations not only allow extension of in vitro life span by overcoming senescence, but also provide a genetic background not unlike that present in cancer precursor lesions because this antigen, by binding to RB and P53, leads to a constitutively active cell cycle in cells deprived of a functional P53 similarly to the situation in serous EUMET precursor lesions. Further characterization of this cell culture model suggested that accumulation of chromosomal abnormalities that are initially left unchecked due to the absence of a normal P53 function eventually overwhelms ability to overcome the spindle assembly checkpoint complex, resulting in a prolonged mitotic arrest. Recovery from such arrest, which is facilitated by lowering BRCA1 expression levels, results in resumption of cell cycle activity without cytokinesis, leading to tetraploidy and, if the process repeats itself, even higher degrees of polyploidy with ensuing aneuploidy ( 77 , 78 ). This scenario is a central element of the model of cancer development proposed in the next section. It is interesting that cultures of extra-uterine Müllerian tumors of low malignant potential derived similarly and in parallel to those of cystadenomas, in spite of absence of normal P53 function due to forced expression of SV40 large T antigen, do not undergo a similar mitotic arrest and that their ploidy status remains substantially more stable than that of cystadenomas ( 79 ). It is tempting to relate such chromosome stability in culture to the fact that these tumors are typically diploid and genetically stable in vivo ( 80 , 81 ). In fact, aneuploid tumors of low malignant potential are associated with a more aggressive clinical course and their response to chemotherapeutic agents may be more typical of carcinomas, raising the possibility that at least some of the those tumors are carcinomas incorrectly diagnosed as LMP tumors ( 81 - 84 ). Indeed, the possibility of using ploidy status as a diagnostic tool to help distinguish tumors of low malignant potential from carcinomas has been suggested ( 80 ). Further understanding of the mechanisms underlying the apparent protection against chromosomal instability in cultures of tumors low malignant potential should further increase our understanding of the development of aneuploidy, one of the hallmarks of cancer.

BRCA1 and BRCA2 mutations, the main determinant of genetic risk for serous extra uterine Müllerian carcinomas, are rare in the sporadic form of this disease. An explanation, which would also account for the site specificity of the tumors that develop in BRCA1 and BRCA2 mutation carriers, is that inactivation of either one of these two genes might mimic important risk factors associated with sporadic extra-uterine Müllerian tumors, and would thus be redundant in individuals exposed to the sporadic risk factors. A possible scenario, given the fact that ovulatory activity is currently regarded as the most important risk factor for the sporadic form of these tumors, would be that a germline BRCA1 mutation could lead to alterations in the dynamics of the menstrual cycle, or in the levels of hormones associated with menstrual cycle progression, that would mimic the consequences of incessant ovulation. The fact that pregnancy or oral contraceptive use, both of which have a strong protective effect against sporadic EUMET, are also protective against these cancers in BRCA1 mutation carriers ( 56 , 57 ) is supportive of this idea of redundancy between genetic and non-genetic risk factors. The finding by Tone et al . ( 58 ) that the expression profile of high-grade serous carcinomas in BRCA mutation carriers resembles more closely that of the fallopian tube epithelium during the secretory phase than the proliferative phase of the menstrual cycle further underscores the role of ovulatory activity associated with menstrual cycle progression in the pathogenesis of BRCA-driven tumors. This hypothesis is referred to as the cell non-autonomous hypothesis of cancer predisposition in BRCA1 mutation carriers. It implies that a germline BRCA1 mutation, in addition to directly influencing cells at risk of cancer development, also influences these cells indirectly and from a distance due to the consequences of such germline mutations on cells involved in menstrual cycle regulation that communicate with tissues at increased cancer risk via either endocrine or paracrine mechanisms. The hypothesis stipulates that the resulting changes in inter-cellular interactions between cells that regulate the menstrual cycle and serous extra-uterine Müllerian epithelium contribute to increased cancer risk in the latter. The hypothesis does not rule out the notion of direct, cell-autonomous effects superimposed on cell non-autonomous effects. The fact that cancerous tissues undergoing chemotherapy in BRCA1 mutation carriers appear to be under selective pressure to regain a normal BRCA1 function suggest a limited cell-autonomous role for BRCA1 mutations once the cancer phenotype has been established ( 59 , 60 ). Chodankar et al . ( 61 ) sought to test the cell non-autonomous hypothesis in an animal model by investigating the consequences of disrupting communications between cells that control the estrous cycle and Müllerian epithelium. Given the central role of ovarian granulosa cells in regulating progression through the normal menstrual cycle, plus the fact that these cells secrete a variety of hormones such as estradiol, Müllerian inhibiting substance, and others that are known to influence EUMET cell growth in vitro , these authors used the Cre-lox system driven by a truncated form of the Fshr promoter to specifically inactivate the Brca1 gene in mouse granulosa cells. The mice indeed developed benign tumors that were clearly of epithelial (as opposed to granulosa cell) origin in strong support of a cell non-autonomous mechanism. These results raise the possibility that EUMET predisposition in human BRCA1 mutation carriers is due, at least in part, to decreased BRCA1 expression in ovarian granulosa cells, thereby disrupting control mechanisms that these cells exert on serous extra-uterine Müllerian epithelium. The findings by Hu et al . ( 62 ) that down-regulation of BRCA1 in primary cultures of human granulosa cells results in increased expression of aromatase, the rate-limiting enzyme in estradiol biosynthesis, is well in line with this hypothesis. It is not clear whether the same mechanism is also responsible for breast cancer predisposition in BRCA1 mutation carriers. The fact that ovulatory activity, which is largely controlled by ovarian granulosa cells, has a strong influence on predisposition to breast cancer in addition to EUMET suggests that the mechanisms of predisposition to breast cancer in mutation carriers could indeed be similar. This idea is further strengthened by the demonstration that oophorectomy can protect against breast cancer in BRCA1 mutation carriers ( 63 ). Hong et al . ( 64 ) sought further insights into the mechanisms whereby a BRCA1 mutation could contribute to cancer development via a cell non-autonomous mechanism. These authors specifically looked at the consequences of a Brca1 mutation in ovarian granulosa cells on the relative lengths of different phases of the estrus cycle and on circulating levels of specific hormones important for the regulation of this cycle. The estrus cycle is equivalent to the human menstrual cycle, the most important risk factor for the sporadic form of serous extra-uterine Müllerian carcinomas. Indeed, the average length of the proestrus phase, which corresponds to the follicular phase of the human menstrual cycle, was longer in mutant mice than in wild type littermates relative to the metestrus phase (corresponding to the luteal phase of the menstrual cycle). Mutant mice also had higher circulating levels of estradiol. They concluded that mice carrying a Brca1 mutation had both increased and prolonged estrogen stimulation unopposed by progesterone ( 64 ). The physiological relevance of these findings is underscored by a follow up study showing evidence of increased steroid hormone stimulation in end-organs targeted by estrogens such as the endometrium and long bones in Brca1 -deficient mice ( 65 ). Mutant mice showed increased endometrial thickness and increased bone length and density, providing strong support for the idea that the presence of a Brca1 mutation in ovarian granulosa cells leads to increased estrogen stimulation. Although the authors used mice carrying homozygous Brca1 mutations in these studies in order to magnify the measurable consequences of such mutations, they also showed that mice carrying heterozygous mutations, such as present in human BRCA1 mutation carriers, showed effects similar to those present in homozygous mutants on sex steroid hormone biosynthesis, albeit of lesser magnitude ( 65 ). Although direct evidence for a cell non-autonomous effect of BRCA1 mutations in human is still lacking, observations with the UK Familial Ovarian Cancer Screening Study cohort (UKFOCSS), which showed that human BRCA1 mutation carriers have a sex hormone dysregulation and altered end-organ hormone-sensitivity during the menstrual cycle, provide strong support for this notion ( 66 ). Endometrial thickness was higher during the follicular phase [OR 1.11, 95% confidence interval (CI), 1.03-1.20; P=0.006) and lesser in the luteal phase (OR 0.90, 95% CI, 0.83-0.98; P=0.027) of BRCA1 mutation carriers compared to non-carriers while median luteal phase progesterone and estradiol levels were respectively 121% (P<0.001) and 33% (P=0.007) higher in mutant compared to controls. These results were not due to differences in oral contraceptive use in carriers ( 66 ).

The cell non-autonomous mechanism of cancer predisposition discussed in this article has important implication for prevention of serous extra-uterine Müllerian carcinomas because it is based on circulating mediators of cancer risk that should be readily targetable with pharmacologic agents or even, perhaps, life style modifications. This underscores the importance of fully elucidating the nature of these mediators and of understanding their consequences on the biology of serous extra-uterine Müllerian epithelium. Better understanding of the cell autonomous mechanisms involved in serous extra-uterine Müllerian carcinoma development, on the other hand, should lead to strategies to reduce the likelihood of recovery from a cell cycle arrest thought to play an important role in the development of these cancers, both as cancer preventive and therapeutic measures. It is hoped that progress in both of these areas will have a significant impact on the morbidity and mortality associated with these important cancers of women.