Endometriosis is the leading cause of morbidity among premenopausal women and the complex pathogenesis of this disease remains controversial despite extensive research. This disease represents one of the most common gynecological problems. It is generally believed that this disease is due primarily to retrograde menstruation or transplantation of shed endometrium. Based on overwhelming data, ovarian endometrioma is considered a neoplastic process, since most endometriosis-associated ovarian carcinoma occur in the presence of atypical ovarian endometriosis. A study comparing patients with typical epithelial ovarian cancer with endometriosis-associated ovarian cancer demonstrated that the patients with the latter disease strongly differ in both biological and histological characteristics. The prevelance of this disease is not completely established, but approximately 15 percent of women suffer from this disease. In addition, we know about the possible links between endometriosis and cancer for almost 100 years. Despite clear evidence revealing that endometriosis increases ovarian cancer risks, it is possible that it may not affect disease progression after the appearance of ovarian cancer. However, despite clear evidence revealing that endometriosis increases ovarian cancer risk, our knowledge of the risk factors is far from established. In our review, we focused on the most recent approaches including possible biomarkers and genetic approaches.

Endometriosis is a common gynecologic disorder usually defined as presence and/or growth of endometrial tissue outside the uterine cavity. Although the prevalence of this disease is not clear, approximately 7%-15% of women of reproductive age[1] have endometriosis, but significantly increases in cases of infertility[2]. Despite the fact that this disorder was first described in 1860[3], the definite cause of endometriosis is still unclear and this disease remains to be one of the most enigmatic female diseases. Probably the most accepted mechanism is the adhesion and subsequent growth of endometrial fragments occurring in the peritoneum by retrograde menstruation. Another possibility is the occurrence of anatomical abnormalities. The repeated events of hemorrhage can also contribute to carcinogenesis and further cancer progression[4]. Lately, the hereditary aspect gained traction, with CYP1A1 m1 polymorphism (particularly CYP10 variable number of tandem repeats allele) and GSTM1 null deletion playing an important role[5]. A detailed summary of possible theories is reviewed by Baldi et al[6] and Yuan et al[7]. In addition, at least three theories supporting the concept of in situ development exist, including a Mullerian and Wolffian rest theory, coelomic metaplasia theory and metaplasia following endometrial stimulation theory (for review see Thomas et al[8]). However, none of these theories explain all the different types of endometriosis[9]. Based on overwhelming data, ovarian endometrioma can be viewed as a neoplastic process.

The possibility of transformation of endometriotic tissue into malignant ovarian cancer has been recognized for almost 100 years[10]. In 1925, Sampson[10] first described malignant changes in endometriosis and proposed the following criteria for diagnosing the carcinomatous development in endometriosis: (1) coexistence of carcinoma and endometriosis within the same ovary, (2) a similar histological pattern and (3) exclusion of a second malignant tumor elsewhere. Later, in 1953, Scott[11] postulated that in addition to the criteria by Sampson, morphological changes demonstrated by benign endometriosis that are contiguous with malignant tissue is a prerequisite for adjudication of a malignancy originating from endometriosis. Histopathological studies have assumed atypical endometriosis to be a transition between benign endometriosis and cancer[12] suggesting the possibility that endometriosis is in fact a premalignant condition. This is supported by the fact that between 60% to 80% of endometriosis-associated ovarian carcinoma occur in the presence of atypical ovarian endometriosis[13].

A summary of more recent cohort studies showed a standardized incidence ratio around 2[14]. Tumors arising from the transformed endometriomas are predominantly of the clear cell and endometroid types[15]. A study of operated ovarian cancer patients showed that full 10% of patients had coexisting ovarian endometriosis. This number increased to 36.8% in patients with clear cell ovarian cancer[16]. These types of studies led to the definition of “endometriosis-associated ovarian carcinoma” (EAOC), which describes ovarian cancer having both cancer cells and endometriosis in the same ovary, presence of cancer in one ovary and endometriosis in second ovary or the presence of ovarian cancer and pelvic endometriosis. Histological evaluations showed significant clinical differences between patients with EAOC and non-endometriosis-associated ovarian cancer[17]. Some studies found that the risk is increased up to 8-fold[18], but the actual increases fluctuate widely among individual studies[19]. For a review of ovarian cancer etiology and risk factors see Hunn et al[20].

However, despite clear evidence revealing that endometriosis increases ovarian cancer risk, the results from meta-analysis suggested that it may not affect disease progression after the appearance of ovarian cancer[21]. In addition, some studies even suggested that the association with endometriosis might have a survival benefit[22]. No associations were observed for common benign gynecologic disorders such as uterine myoma, adenomyosis or endometrial polyps[23]. On the other hand, Brinton et al[24] showed an increased risk of extra-pelvic carcinomas (including breast cancer) in patients with endometriosis. The newest results are less clear since the outcome of studies examining the association between endometriosis and breast cancer are often contradictory[14] and the newest population-based cohort studies show no overall association[25]. A study evaluating possible risk factors revealed only one-ovarian endometriomas of 9 cm or more in diameter was a strong predictor of development of ovarian cancer[26].

An interesting observation was the measurement of the incidence of ovarian cancer confined to endometrial polyps, which can reach between 0.5% to almost 5% depending on the diagnostic methods[27]. These polyps can be found with increased frequency in women with endometriosis, but the retrospective analysis did not reach clear conclusions[28].

Despite the extensive research, the exact mechanisms leading to malignant transformation of endometrial tissue are not established. In endometriosis-associated cancer, the benign-appearing ovarian masses can be detected several years before cancer diagnosis[29].

Ovarian cancer resulting from endometriosis has some special characteristics such as having endometroid or clear cell histology, being diagnosed earlier than other types of ovarian cancer and have a better prognosis[30]. However, the question of endometriosis being a prognostic factor for cancer survival is not clear. On one hand, some studies found no definitive association between the presence of endometriosis and survival[31]. On the other hand, a large Swedish study found significantly better survival in women with endometriosis than for all malignancies combined. In the cases of breast and ovarian cancer, this survival was even more pronounced[32]. The protective effect for both one-sided oophorectomy and radical excision of all visible endometriosis was described[33]. The most recent meta-analysis of risks and prognosis of ovarian cancer confirmed that endometriosis is strongly associated with the increased risk of ovarian cancer and these cancers show favorable characteristics including low grades and specific histology[21].

Both diseases are progressive and estrogen dependent, often associated with late menopause and infertility. In addition, tubal ligation, hysterectomy and progesterone treatment can decrease the incidence of both diseases[34].

Despite the fact that postmenopausal endometriosis is rare, there is a risk in patients taking hormone therapy[35]. Hormone therapy, particularly estrogen therapy, is often supposed to stimulate the growth of endometriosis, with additional increase in the risk of ovarian cancer development, which is particularly dangerous in postmenopausal endometriosis, where estrogen therapy should be used only in combination with progesterone[35]. Estrogen-induced triggering is the same in endometriosis and estrogen-dependent cancer. The quantification of estrogen derivates was even suggested to be a prognostic factor in ovarian cancer development[36]. In vitro experiments using ovarian cancer cell lines showed similarly increased expression of ER-α receptors as in active endometriosis[37] which was connected with the pathogenesis of endometriosis. Implants of endometriosis contain estrogen, progesterone and androgen receptors, but the effects of these hormones are opposite-estrogen stimulates proliferation and androgens cause atrophy and regression. The findings show that the basalis has approximately five times as many lymphatic vessels as the functionalis, but it cannot be currently explained[38].

Since several theories about the causes of endometriosis failed to explain all types of lesions and suggest the possibility of combined mechanisms, recent hypothesis include the likelihood that endometriotic lesions arise from ectopic endometrial stem cell progenitors (for review see Maruyama et al[39]). The origin of oocytes in adult ovaries has been disputed for a long time. Some of the recent studies suggested the possibility that newly implantation of persistent fetal stem cells might play role in tissue regeneration and/or growth. A finding that several HOX genes controlling lineage infidelity in ovarian cancer[40] are expressed in primitive hematopoietic cells suggested a role in early hematopoietic differentiation. Some studies even suggested that HOX genes expressed in ovarian epithelial cells might regulate cancer stem cells. At the same time, stem cell transformation can be underlying cause of ovarian cancer[41,42].

The main risk factor is clearly the age at endometrioma diagnosis. As independent prognostic factors, endometrioma of a diameter above 9 cm and postmenopausal status were established[26]. There is some evidence that drugs employed in in vitro fertilization might increase the risk of ovarian cancer[43]. However, some studies found opposite results, so at present, no clear conclusion can be reached.

An effort to determine the potential association between endometriosis and ovarian cancer led to the study evaluating the expression of possible biomarkers which have been previously shown to participate in the pathogenesis of these diseases. The results revealed that endometriosis and epithelial ovarian cancer cells manifested significantly higher levels of mRNAs of transforming growth factor-b1, cyclooxygenase-2, vascular endothelial growth factor, ER-1a, aromatase and androgen receptors, whereas the mRNA levels of progesterone receptors were much lower[44]. The clinical importance of this report is somewhat limited due to the low number of patients; however, its conclusion deserves further investigation.

An additional possible biochemical marker might be cancer antigen CA125[13], which is commonly used for the monitoring of epithelial ovarian cancer. Since endometriosis is also often associated with a high level of this antigen, it cannot be used for differential diagnosis. A detailed study suggested that a CA72-4 antigen can serve in this role as a biomarker useful to confirm the benign nature of endometriomas in patients with high CA125 levels[45].

An immunohistochemistry study of ovarian cancers arising from endometriosis confirmed that the estrogen-dependent cancers are substantially more associated with endometrioid adenocarcinoma than clear cell carcinomas. The positive hepatocyte nuclear factor-1 beta was common in clear cell carcinomas, but rare in endometrioid adenocarcinoma, which correlated with p53 staining, but reversely correlated with estrogen receptor presence[46].

Suryawanshi et al[47] focused on plasma microRNAs as a novel biomarker. Using a reverse transcriptase quantitative polymerase chain reaction, the authors identified 23 microRNAs which are differentially expressed in healthy people and patients with either endometriosis or EAOC. These microRNAs were subsequently further evaluated in a larger cohort. The results showed that plasma microRNA expressing patterns might serve as specific and reliable diagnostic biomarkers[47] resulting in some authors suggesting than microRNA studies will lead to changes in current treatment of both endometriosis and ovarian cancer[48].

A wide variety of molecular alterations have been reported to be involved in the malignant transformation of endometriosis, some of which are common in both endometriosis and ovarian cancer, whereas the others are universal among various tumors. Only a very few were suggested to specifically refer to the malignancies of endometriosis.

Many hypothesize that changes in the expression of tumor suppressor genes and oncogenes occurring in the eutopic endometrium might lead to overgrowth of endometrial foci outside the uterus[49,50].

Microsatellite analysis showed that loss of heterozygosity on p16 (Ink4), gut-associated lymphatic tissue (GALT), and p53 occurs in endometriosis. Similarly, activation of mutated K-ras gene is an important step in both genesis and progression of ovarian cancer. Alternation of p53 gene caused aberrant regulation of the H-ras protooncogene[51]. Using a murine model of mutationally activated K-ras gene, it was demonstrated that these animals develop both endometrial lesions and the ovarian tissue. The following mutation blocking the expression of PTEN caused ovarian cancer[52]. K-ras mutation may promote carcinogenesis of endometriosis leading to ovarian clear cell carcinoma[53].

Mutations in ARID1A and PIK3CA were first described in numerous cases of ovarian clear cell carcinoma, but later also in precursor endometriosis tissues. In addition, PTEN-PIK3CA-mTOR pathway was strongly implicated by finding PIK3CA mutation in up to 46% clear cell ovarian cancer[54]. PIK3CA mutation is considered to be an early event in the development of endometriosis-associated ovarian clear cell adenocarcinoma. For a better understanding of the effects of these two mutations, functional studies evaluating effects of individual mutations separately and in both normal endometrium and endometriotic epithelium are necessary. The sequence of events leading from normal eutopic endometrium to endometriosis and subsequent ovarian cancer is still hypothetic[55].

Some studies suggested that a histologically normal endometrium may bear genetic damage caused by iron-dependent oxidative stress[56]. Some authors suggested that suppression of pre-apoptotic gene Bax and/or upregulation of anti-apoptotic gene Bcl-2 can be involved in endometriosis and malignancies[57].

Genetic instability might lead to both endometriosis and ovarian cancer. It can include deactivation of some tumor-suppressing genes, changes in activity of enzymes involved in DNA repairs or mutations in genes such as GALT and GSTM. Similarly, mutations in tumor suppressive gene PTEN was often found both in endometric and cancer tissues[58]. In addition, c-erbB-2 and p53 genes have been found to associate with endometriosis-related ovarian cancer[59]. The recent study showed that endometriotic lesions have mutations in cancer-related genes such as PTEN, KRAS, p53, and ARIDIA[48,60].

In addition, both endometriosis and EAOCs share some of the mediators implicated in inflammatory angiogenesis. They exhibit genetic polymorphisms of several genes including intercellular cell adhesion molecule-1, interleukin (IL)-6 and IL-10 promoters, tumor necrosis factor-alpha, and nuclear transcription factor-κB[61]. Further genetic factors such as loss of heterozygosity, K-ras, P53, and PTEN mutations or hepatocyte nuclear factor-1β were suggested and tested (for review see Kobayashi et al[4]). Genome-wide association studies and as well as transcriptome sequencing have shown that genes from the 1p36 region might be important in both endometriosis and endometriosis-associated cancer development[62]. In a recent study, the endometriosis-associated ovarian carcinogenesis has been linked to oxidative stress-induced increased genomic instability, aberrant methylation, and aberrant chromatin remodeling, as well as mutations of tumor suppressor genes[63]. For a summary of the molecular biology aspects of ovarian cancer in endometriosis, see Mandai et al[64].

Endometriosis is a multifactorial disease and despite intensive research in the last several decades, many questions remain to be answered as to the exact events leading from endometriotic cysts to endometriosis-associated ovarian cancer. The probable mechanisms and attributing factors include longstanding estrogen stimulation, repeated heavy menstruation resulting in tissue damage and early events on the molecular level were repeatedly shown. Numerous markers have been proven to correspond with this type of cancer. Thus far, however, no single one can be used for diagnosis, not to mention treatment. Despite the fact that the links between endometriosis and ovarian cancer appear to be clear, we have to keep in mind that having endometriosis might be less risky than undergoing in vitro fertilization, which can increase the risk of ovarian cancer three times[65]. The exact mechanisms of the endometriosis-ovarian cancer conversion are still not fully established and the need for new approaches in the understanding and treatment of endometriosis is urgent.