The Association Between Endometriosis and Surface Epithelial Ovarian Tumors: A Review of Pathological Data

The aim of this study was to find the proportions of ovarian epithelial tumors, which are associated or concurred with endometriosis and to evaluate whether these tumors are originated from endometriosis or not. Endometriosis-associated ovarian tumors have been investigated and described in several studies ( 15). The most common histological subtypes of ovarian epithelial tumors found to be associated with endometriosis are endometrioid and clear cell subtypes ( 15, 17). We reviewed all diagnosed surface epithelial ovarian tumors including carcinoma and borderline tumors in a 7-year period. Out of 182 cases, only 32 (17.6%) had concomitant endometriosis, which gave the impression to be similar that reported by other studies. As reported in an Iranian study, the prevalence of endometriosis was 38% in infertile and 11.6% in fertile females. There are many pitfalls regarding the diagnosis of endometriosis in patients with ovarian tumors, leading to misestimating the prevalence of endometriosis in such patients. Firstly, there are numerous criteria for the diagnosis of endometriosis-associated ovarian tumors and also detection of a small focus of endometriosis demands extensive sampling of the specimen, which is not always possible. Also in cases that the tumor is adjacent to endometriosis, the main problem is the destruction of the endometriotic foci by tumor and the detection of endometriosis is not always possible and, finally, sometimes the endometriotic site is marsupialized or never treats during the surgery causing underestimating endometriosis in such cases ( 12, 18). The association between ovarian tumor and endometriosis was reported in the literature by 2 different categories. A transition point from an endometriotic lesion (atypical endometriosis) to a frankly ovarian invasive carcinoma has been demonstrated in 36% to 42% of cases ( 19- 24). In the remaining cases, the tumor simply coexists with endometriosis without a transition point. Atypical endometriosis has been known to be premalignant lesion of ovarian carcinoma. In 1979, Morris et al. categorized atypical endometriosis according to histomorphological features from mild to severe. Endometriotic foci with mono-layered epithelium consisting of eosinophilic cuboidal cells with pleomorphic enlarged nuclei are considered mild atypical endometriosis and those cases with tufting, crowding, stratification, and intraluminal projections of cells with eosinophilic cytoplasm and large hyperchromatic pleomorphic nuclei are categorized as severe atypical endometriosis. Glandular proliferation with mild atypia can be a reactive condition due to local inflammation and it should not be considered as a premalignant lesion, on the other hand, severe atypia is only expected in potentially premalignant lesions ( 12). In a study conducted by Fukunaga et al. the prevalence of endometriosis-associated ovarian cancer was 24%, of which 61% were accompanied with atypical endometriosis and in our study, 12.5% (4 out of 32 cases) of patients with endometriosis-associated surface epithelial tumor had atypical endometriotic changes in the background, all of which were endometrioid carcinoma. In the remaining 28 cases, endometriotic foci could not fulfill the criteria for atypical endometriosis. As a result, we concluded according to Van Gorp et al. the occurrence of endometriosis-associated ovarian tumors is minimum considering only definite cases with a proven atypical transition (Category A) and it is maximum when other cases with separated foci of endometriosis (Category B and C) are included. However, in either way, the estimated prevalence of endometriosis-associated ovarian epithelial tumors is lower than the actual prevalence. The reason of the mentioned problem is scarcity of the examined specimen and further missing of the transitional area, and destruction of premalignant endometriotic foci in some cases. This observation in a surgical specimen should aware the pathologist to examine the entire lining of the cyst such as surface epithelial ovarian borderline tumors and the patient should be under close follow up. The histological subtypes of clear-cell and endometrioid ovarian cancer are four-fold higher in patients with endometriosis; therefore, these subtypes are recently regarded as “Endometriosis-associated ovarian tumors” ( 6). Another study described a two-fold increase of low grade serous carcinomas in patients with endometriosis, but mucinous or high-grade serous carcinomas were not related with endometriosis in other studies ( 25). According to the results of this study, only 1 low-grade serous carcinoma (3.12%) associated with endometriosis was not found with high-grade serous carcinoma. If we define ovarian cancer as type I and type II according to Kurman and Shih’s classification ( 26), coexisting endometriosis was more commonly detected in type I cancers. It should be mentioned that endometrioid, clear cell and low-grade serous carcinoma are considered as type I cancer. Consistently, in a study carried out by Wang et al., 18.3% of endometriosis-associated ovarian tumors were type I ovarian carcinoma and endometriosis was postulated to be one of precursor lesions for type 1 ovarian carcinoma ( 27). These studies give some clues for discovering pathogenesis mechanism of type 1 ovarian carcinoma and emphasizing the difference between type 1 (low-grade serous) and type 2 (high-grade serous) ovarian carcinomas. The mean age of our patients was 41.6 years that was less than patients without endometriosis, (43 years). In a study conducted by Aris et al. the mean age of women was 48.3 years, which is higher than the value in this study ( 20). In Aris’s study, there was a statistically significant difference between the mean age of patients with ovarian cancer, which is not the case for endometriosis (53.9 years) and ovarian cancer with endometriosis (48.3 years) (P = 0.003), suggesting an early onset of ovarian cancer in women having endometriosis ( 17, 18), but we have not found statistically significant differences in the present study (P = 0.57). Also we have not found statistically significant differences between various ages and concomitant endometriosis with endometrioid and clear cell carcinoma, when age was divided into decades, a statistically significant difference was obtained in age 40 to 50 years (P = 0.034). In the present study, out of 32 endometriosis-associated surface epithelial ovarian tumors, the endometriosis was found: 25 cases were in the same side, 4 cases bilateral, 1 case in opposite ovary, 1 case within fallopian tube, and the last one in abdominal wall. These distributions show that malignancies were more common in ovaries that contained endometriosis (78%) than in the opposite ovary or extra ovarian locations of endometriosis. It seems that some local factors have a significant role in the transformation pathway from endometriosis toward malignancy. The findings of this study revealed only a single case of extra-ovarian endometriosis-associated clear cell carcinoma (in abdominal wall related to previous cesarean scar). In ovarian endometriosis, endometrioid adenocarcinomas accounted for 69% of tumors, clear cell carcinoma 14%, and sarcomas 12% ( 4). In contrast, clear cell carcinoma and adenosarcoma were more associated with extraovarian endometriosis ( 4). We have found that 2 out of 15 cases of concomitant endometriosis with endometrioid carcinoma of ovary have shown endometrioid carcinoma in the uterine cavity (13.33%). Concurrent primary neoplasms in female genital tract are a matter of interest for clinicians. Estrogen receptors may play a rule in synchronous or multifocal malignancies in this region. Endometriosis has been shown to be associated with such situations. Thus, these finding show that ectopic endometrium as well as normal endometrium can be stimulated by estrogen hormone ( 10). Data regarding features and classification of endometriosis-associated ovarian epithelial tumors in this study are illustrated in Table 2. We concluded that the actual prevalence of endometriosis-associated ovarian tumors is underestimated. One reason is the lack of enough confident criteria for the diagnosis of endometriosis-associated ovarian tumors and also extensive sampling of surgical specimen for detecting small or destructed foci of endometriosis is not always possible.