Ovarian cancer in endometriosis: clinical and molecular aspects.

Epithelial ovarian cancer can be divided into two subtypes based on clinical, histologic and molecular factors: Type I and Type II. 13 Type I tumors are slow growing and include low-grade serous carcinoma, mucinous carcinoma, endometrioid carcinoma, malignant Brenner tumor, and clear cell carcinomas. In contrast, Type II tumors are high-grade and include high-grade serous carcinoma (moderately and poorly differentiated), malignant mixed mesodermal tumors (carcinosarcomas) and undifferentiated carcinoma. 13 FIGO stage and grade of ovarian cancer are significantly correlated with five-year survival rates in ovarian cancer, with earlier stage and lower grade predicting better outcomes. 14 , 15 Research has attempted to determine if unique predictive factors exist for EAOCs. EAOCs represent a subclass of ovarian neoplasms that have specific clinical characteristics that include histology, FIGO stage, CA125 levels, patient age, menopausal status at diagnosis and survival outcomes. 16 Studies suggest that the majority of EAOCs are endometrioid and clear cell subtypes, with endometriosis found in 30–55% of clear cell and 30–40% of endometrioid ovarian cancers. 5 , 17 – 20 There is strong evidence to suggest EAOCs are more likely to be diagnosed at both an earlier stage and lower grade. Wang et al . stated that 88.2% of EAOCs were diagnosed at Stage I versus 15.8% of non-EAOCs. Kumar et al . found 49% to be FIGO I or II; and Erzen et al . reported that 67% of EAOCs were diagnosed at Stage I versus 27.6% for non-EAOCs. Similarly, grade I/II tumors were found more commonly in patients who had EAOCs than in those who had other ovarian tumors. 18 , 21 Though most studies have focused on outcomes of EAOCs compared to all other ovarian cancers combined, Davis et al . have recently compared clinical outcomes of the most common type of ovarian tumor, papillary serous, and outcomes of EAOCs. The results showed that compared to papillary serous subtypes, EAOCs were associated with an earlier stage (I/II), lower grade and younger age at time of diagnosis, as well as a lower rate of recurrence. This is consistent with other articles in the literature including one recent study that found patients with EAOCs were six years younger and 35% more likely to be premenopausal than women with other ovarian cancers. 17 , 18 , 22 , 23 This notion was further supported by Mangili et al ., who found that the average age at diagnosis for women with EAOCs was 55 years while the average age of women diagnosed with non- EAOCs was 62 years (P=0.03). As reviewed above, most studies confirm that EAOCs tend to be associated with younger ages, earlier stages and lower grades at diagnosis. Despite these seemingly beneficial prognostic factors, research has not demonstrated a difference in five-year survival rates between patients with EAOCs and non-EAOCs when stage was controlled. 18 , 19 , 21 , 24 Because earlier stage at diagnosis is associated with a better prognosis, evaluation of potential tumor markers to screen for malignancy is a constant area of research. CA125 is a marker associated with ovarian cancer that has a high sensitivity but poor specificity for diagnosis of disease; however it is considered part of the standard of care for surveillance of ovarian tumors. Its relationship in EAOCs has been studied and results are controversial. A recent study by Wang et al ., found that patients with EAOCs had significantly lower levels of pre-surgical CA125 (122.9 vs . 1377.5). However, most other studies have found no significant difference between patients with EAOCs and non-EAOCs. 17 , 21 , 23 , 24 One area that is unique to EAOC tumors is the finding of concurrent malignancies at time of diagnosis. In up to 10% of women diagnosed with EAOCS, a secondary cancer diagnosis has been made concurrently. 25 Davis et al . found 23.8% of EAOCs had a concurrent primary cancer diagnosis of which 94.1% were endometrial cancer. This finding was also replicated by Mangili et al . who reported that 40% of patients with EAOCs also had a diagnosis of endometrial cancer. Furthermore, of these patients 33% of them had an endometrioid EAOC subtype and 94% had the same histology in both the ovarian and uterine malignancy. This relationship between EAOCs and endometrial cancer has led to much debate over the molecular mechanisms that dictate transformation of endometriosis to ovarian cancer.

As discussed above, endometrioid and clear cell (EAOCs) are thought to develop from endometriosis and have distinct mutations in genes and pathways that distinguish them on a molecular level from non-EAOCs. New discoveries are constantly being made, but the most commonly associated mutations include a loss of heterozygosity in PTEN mutations (20%), beta-catenin gene mutations (16–54%), KRAS mutations (4–5%), microsatellite instability (13–50%) and ARID1A mutations (40–50%). 26 , 27 In 2000, Sato et al . were the first to identify that mutations leading to inactivation and loss of heterozygosity (LOH) of the tumor suppressor gene PTEN (locus 10q23.3) were associated with both endometrioid and clear cell carcinomas in endometrial and ovarian cancers. Subsequent research identified inactivation of PTEN as an early event in the malignant transformation of endometriosis to EAOC, which is thought to be involved in the development of 14–20% of epithelial ovarian cancers. 28 , 29 Characterization of the PTEN pathway found that the PTEN mutation is involved in the PI3K signaling pathway with LOH resulting in activation of PI3K leading to development of clear cell subtypes of EAOCs. 30 , 31 Supporting evidence for PTEN involvement in this malignant transformation includes both mutations of PTEN found in benign ovarian endometrioid cysts as well as studies that identify PTEN mutations in patients with ovarian cancer and adjacent endometriosis. 12 , 31 Mouse models have also shown concomitant PI3K/PTEN and Wnt signaling pathways are sufficient to induce EAOCs. 32 Recent research by Govatati et al . 2013, may have further characterized the transformation of endometriosis to EAOCs by demonstrating increased mutations in PI3K/PTEN and the downstream target, Akt, in eutopic endometrial tissue of patients with endometriosis. This suggests that the initial mutations occur in eutopic endometrial cells that subsequently migrate out to become ectopic tissue (endometriosis). Studies into the characterization of the downstream effects of PTEN mutations are ongoing as exciting clinical advances suggest PTEN overexpression can enhance the anti-tumor effects and even reverse the resistance of some chemotherapeutic drugs. 33 , 34 TP53 (locus 17p13.1) is a tumor suppressor gene that encodes for nuclear protein p53. Mutations are related to an over-expression of the non-functional form that causes an accumulation of protein in the nuclei of cells. 35 , 36 It has been identified as an important precursor in ovarian cancer, but controversy exists over its involvement in type I ovarian tumors. Early studies examining the role of TP53 in transformation of endometriosis to carcinoma found mutations to be associated with severe/late stage endometriosis. Similar studies found p53 protein accumulation in endometriosis adjacent to clear cell and endometrioid carcinoma and one study reported a statistically significant difference in rates of p53 expression in the transition from typical to atypical to ovarian cancer associated with endometriosis. 7 , 37 , 38 However, many of these studies had very small sample sizes and did not demonstrate statistically significant differences in these associations. Some studies suggest that as few as 5% of type I tumors harbor TP53 mutations. 35 More recently, it has been suggested that TP53 mutations exist in most of the Type II tumors, but only in advanced type I ovarian tumors. TP53 is now most commonly associated with high grade serous ovarian cancer and advanced ovarian cancers, with 92–96% of these tumors expressing a mutation in this tumor suppressor gene. 39 , 40 With a growing body of evidence supporting that TP53 mutations are more often associated with non-EAOCs, there has been a decrease in focus on TP53 and its involvement in the transition from endometriosis to EAOCs. Kras is an oncogene that encodes a GT-Pase transductor protein involved in regulating cell division by relaying external signals to the cell nucleus. Kras mutations impair the ability of it to switch between active and inactive forms leading to dysregulated cell proliferation. 41 Mutations occur in 3.7–36.4% of endometrioid cancers, with most studies suggesting a mutation rate of approximately 10%. 29 In 2005, a mouse model comparing activation of Kras in endometrial cells deposited in the peritoneum to activation of Kras in peritoneal cells showed that the endometrial cells progressed not only to implants, but Kras and PTEN mutations concurrently progressed to endometrioid cancer. 29 , 40 However, recent evidence suggests that although mutations in Kras occur in both endometrioid and clear cell ovarian tumors, they appear to be most prevalent in mucinous, low grade and well-differentiated ovarian cancers. 42 , 43 Further studies are warranted to determine whether prognosis is impacted by the presence of Kras, as some research has suggested that survival may be increased in patients with this mutation. 43 As technology has advanced, the capacity to perform genome-wide analyses and the opportunity for whole exome and RNA sequencing to assess somatic mutations in ovarian cancer has developed. In 2010, two independent studies revealed that somatic mutations in ARID1A were major molecular contributors to clear cell and endometrioid ovarian cancers. 27 , 44 The majority of ARID1A mutations are either frame shift or nonsense suggesting its role as a tumor suppressor gene and the protein it encodes, BAF250a, is part of a multiprotein SWI/SNF chromatin remodeling complex involved in the regulation of cellular processes including differentiation, proliferation, DNA-repair and tumor suppression. 45 , 46 In a study of EAOCs, Wiegand et al . found 73% of clear cell and 50% of endometrioid ovarian cancers with an ARID1A mutation to have a loss of expression of BAF250a compared to approximately 10% of non-EAOCs. They confirmed this relationship with immunohistochemical studies where silencing of ARID1A resulted in a loss of BAF250a expression. 27 Subsequent research has shown the ARID1A mutation to be present in approximately 50% of clear cell and 40% of endometrioid ovarian cancers. 44 , 47 Further investigation has found specimens where ARID1A was mutated in both the ovarian tumor and adjacent endometriosis providing support for its involvement in the progression of endometriosis to carcinoma. Loss of ARID1A has also been demonstrated in 26% of uterine endometrioid carcinomas and has been found to correlate with uterine endometrioid tumor progression from low grade to high grade. 46 , 48 Though global study of different neoplasms has noted somatic ARID1A mutations are present in gynecological cancers, it is appears to be isolated to EAOCs and endometrioid uterine cancers. 49 , 50 Current research is focused on correlating ARID1A mutations with other mutations in the EAOCs and its potential for use as a prognostic factor for malignant neoplasms. Huang et al . recently reported that ARID1A mutations coincide with activation of the PI3Kc-AKT pathway that has been identified in ovarian clear cell carcinomas. 51 Though loss of ARID1A has not yet been identified as a prognostic factor in ovarian cancer, preliminary research in breast cancer has suggested that absence of BAF250a is associated with poorer postoperative disease survival and suggests that it may be useful as a target for breast cancer treatment. 52 CTNNB1 is the gene that encodes the β-catenin protein that is involved in the Wnt/β-catenin pathway and regulation of cellular processes including proliferation, motility and survival. 32 It is found in up to 40% of endometrioid ovarian carcinomas and both accumulation and depletion of β-catenin may be associated with dysregulation leading to malignancy. 53 , 54 CTNNB1 has been associated with squamous differentiation, low tumor grade and good prognosis in endometrioid ovarian cancers. 39 , 53 Recent in vivo murine models have shown that alterations in the Wnt/β-catenin signaling pathway promote eutopic endometrial cell invasion and adhesion providing evidence for the involvement of this pathway in transformation of endometriosis to ovarian cancer. 55 Another study investigated the effects of the signaling pathway in the mechanism of endometriosis associated fibrosis and reported that activation of the Wnt/β-catenin by treatment with recombinant Wnt lead to a prevention of fibrosis associated with endometriosis. 56 Microsatellites or MicroRNAs (miRNAs) are endogenous 22 nucleotide-long, highly conserved, noncoding RNAs thought to be primarily involved in the negative regulation of target genes in cell cycle progression, proliferation and differentiation. 57 Microsatellite instability is defined as a size change in short tandem repeat sequences in a tumor compared to normal tissue of the same patient and has been implicated in both 7% to 22% of sporadic ovarian cancers and is commonly associated with BRCA1 or BRCA2 germline mutations of inherited ovarian cancers. 58 , 59 Studies on miRNA expression have shown down regulation of specific miRNAs in clear cell and endometrioid ovarian cancers, decreased expression in recurrent versus primary ovarian cancers and up to 25 fold reductions in moderate to severe endometriosis when compared to normal endometrium. 57 , 60 , 61 MiRNA profiling in the plasma of healthy individuals compared to individuals with endometriosis and ovarian cancer found novel miRNA markers that could distinguish between these groups with good sensitivity and specificity. 57 , 62 However, studies comparing expression of high, low or stable microsatellite instability have failed to find any differences that can distinguish between specific histologic subtypes of ovarian cancers or provide prognostic value. 63 , 64

The advancement of technology has led to discoveries surrounding the genetic and molecular makeup of endometriosis and ovarian cancer both as independent entities and as EAOCs. As knowledge has expanded, evidence points to EAOCs as malignancies that develop due to mutations in tumor suppressors and signaling pathways involved in the regulation of the cell cycle and proliferation Distinct molecular profiles of high grade, more advanced cancers and low grade, less aggressive tumors in ovarian malignancies are being developed. While Type I tumors are slow growing, tend to present confined to the ovary in early stages, and are associated with mutations such as KRAS, CTNNB1, PTEN, PIK3CA and ARID1A, Type II tumors are rapidly growing therefore presenting at later stages and are associated with TP53 and BRCA mutations. 39 Furthermore, it has been suggested that within Type I tumors EAOCs (clear cell ovarian and endometrioid ovarian carcinoma) there are very distinct profiles as indicated by frequency of mutations associated with each of these neoplasms. Two predominant theories surrounding the development of EAOCs have developed as this evidence unfolds. One theory suggests that invasive endometriosis leading to ovarian carcinoma results from phenotypically normal endometrial cells that are displaced into the pelvis by retrograde menstrual flow and develop the adhesive and proliferative properties of endometriosis. Over time, these cells further mutate to become invasive and subsequently lead to ovarian cancer. Another theory proposes that the inciting event is the mutation in eutopic endometrium that allows for endometrial cell migration through the fallopian tubes and proliferation outside of the uterus resulting in an invasive endometriosis phenotype and subsequent carcinoma. This theory is supported by molecular differences originally thought to be isolated to endometriotic tissue, but were later found to also be present in eutopic endometrial tissue of women with endometriosis and absent in endometrial tissue of disease free women. 1 , 65 Because there is still much to be learned about EAOCs, continued efforts on early detection through close monitoring of patients symptoms should be emphasized. When women with a history of endometriosis become menopausal, general practitioners and gynecologists should have a heightened suspicion for EAOCs. Endometriosis is typically diagnosed two to three years earlier than ovarian cancer; therefore, a change or increase in symptoms warrants appropriate evaluation. 66 Further defining the specific characteristics of EAOCs including younger age, lower tumor stage and grade at diagnosis has allowed providers to be able to better detect these tumors prior to metastases. However, if these tumors spread outside of the ovary, they do not respond well to chemotherapy. Therefore in addition to monitoring patient symptoms, prevention efforts should also be instituted. Good evidence exists that reducing the number of ovulations in a women’s life has a significant effect on reduction of ovarian cancer and research shows that oral contraceptive use for 5 or more years reduces the risk of epithelial ovarian cancer by as much as 50%. 67 Similarly, the effect of tubal ligation as a protective mechanism in the development of all ovarian cancers by preventing the migration of endometrial cells out of the uterus has shown a 30% risk reduction in the development of ovarian malignancy. A large meta-analysis by Sieh et al ., specifically demonstrated a 50% risk reduction in both clear cell and endometrioid ovarian cancers with tubal ligation. 68 Thus, careful contraceptive counseling by providers can help to reduce the incidence of ovarian cancers by preventative methods. With continued efforts and future discoveries, the hope is to eventually develop screening tests and more sensitive detection methods leading to better prognoses and reduced mortality from ovarian cancer. As we move towards this goal, caregivers in the clinical setting are the first line defense, providing appropriate counseling and careful attention to patient concerns in an effort to reduce development and increase early detection of ovarian cancer.