Main

Epithelial ovarian cancer has the highest mortality rate of gynecologic malignancies. It was diagnosed with an incidence of 1400 women in The Netherlands in 2010, whereas 1065 women died from this disease.1, 2 The vast majority of epithelial ovarian cancers are high-grade serous carcinomas, accounting for 70% of cases. High-grade serous carcinomas are mostly diagnosed in an advanced stage of disease due to non-specific symptoms and a tendency for early peritoneal spread.3 The overall survival of patients with ovarian carcinoma has improved only minimally, and one of the reasons is that its route of development has not been totally clarified yet.

It is thought that ovarian carcinoma develops from tissues that are embryologically derived from the Müllerian ducts and that the ovaries are involved secondarily.4, 5 Müllerian ducts develop in early embryogenesis into the upper part of the vagina, uterus, and fallopian tubes.5 In the tubes the in situ carcinoma, serous tubal intraepithelial carcinoma, has been identified.6, 7 Serous tubal intraepithelial carcinoma is found in 41% of patients diagnosed with serous ovarian carcinoma.6, 8, 9, 10 However, in a significant part of serous ovarian cancer patients, no serous tubal intraepithelial carcinoma or predisposition for serous carcinoma has been found. Recently, endometrial intraepithelial carcinoma, was suggested as an alternative source for serous ovarian cancers, but its prevalence in ovarian cancer patients is unknown.11

The standard surgical treatment of serous ovarian carcinomas includes bilateral salpingo-oophorectomy and hysterectomy. The most important reason to perform a hysterectomy is that the uterine serosa is one of the most frequent sites of peritoneal spread.12 Moreover, double primary endometrial carcinomas are identified in 10% of endometrioid ovarian carcinomas and in 2–5% of other ovarian carcinoma types.13, 14, 15

The current study was performed to analyze the cervix, endometrium and tubal epithelium thoroughly in serous ovarian cancer patients. Therefore, structural embedding of the entire endometrium and fallopian tubal epithelium was performed and extensive histological assessment of tissue for all cases.

Materials and methods

Study Design and Case Selection

In this prospective, multicentre, cohort study, all women with high-grade serous ovarian cancer undergoing a primary or interval debulking surgery that included hysterectomy and bilateral salpingo-oophorectomy were identified. Women were operated at the Radboud University Nijmegen Medical Centre and Canisius Wilhelmina Hospital, Nijmegen, from September 2010 to December 2012. Surgery was performed by a gynecologic oncologist. Clinical parameters, such as age, menopausal status, body mass index, BRCA mutation status, treatment protocol, and medical history, were retrieved from medical records and pathology reports. Patients were classified and staged in accordance with the WHO classification and FIGO surgical staging adopted in 1988.16, 17 Carcinomas were classified as primary ovarian if the largest tumor bulk was located in the ovaries. In case of simultaneous carcinoma involvement of the endometrium, additionally depth of myometrial invasion, ovarian tumor pattern, lymfovascular invasion and presence of serous tubal intraepithelial carcinoma or endometrial intraepithelial carcinoma were used to differentiate primary ovarian from endometrial carcinoma.18

Histological Examination of Tissues Derived from the Müllerian Duct: Endocervix, Endometrium and Fallopian Tubes

All tissues were macroscopically assessed and described before formalin fixation overnight. The anterior side of the uterus was once cut several centimeters in longitudinal direction, from cervix toward fundus, to enable optimal fixation. The uterus was sampled using a systematic protocol with complete embedding of the endometrium, the SEE-End (Sectioning and Extensively Examining of the Endometrium) protocol (Figure 1). First of all, the uterus was cut longitudinally into two equal parts. Routine sections of the cervix uteri anterior and posterior were embedded, followed by a section of the isthmus region. A representative section of the fundus region was cut in longitudinal direction. Remaining endometrium was sectioned in transverse direction at intervals of 2–3 mm. For practical reasons, redundant myometrium was excised from additionally sampled sections to enable embedding of multiple endometrial samples in one paraffin block (Figure 1). In case of an endometrial polyp, it was embedded in toto. Representative sections were embedded from polyps of the cervix or isthmus, leiomyomas, or other lesions for histological examination.

Figure 1
figure 1

The entire endometrium was sampled and embedded following the SEE-End protocol (Sectioning and Extensively Examining the Endometrium). After sectioning the standard diagnostic samples, the remaining endometrium was sectioned in transverse direction with intervals of 2–3 mm. Redundant myometrium was removed from additionally sampled endometrium to enable the display of several endometrial sections in one slide. (a) Cervix uteri anterior and posterior, (b) Isthmus, (c) representative section of fundus region of the endometrium, including an endometrial polyp. (d) Representative slide of small leiomyoma, (e–h) remaining endometrium and endometrial polyp were embedded in accordance with the SEE-End protocol.

Full size image

The fallopian tubes were completely embedded for histological examination, in accordance with the SEE-Fim protocol (Sectioning and Extensively Examining of the Fimbriated end).19 The fimbriated end was amputated at the infundibulum and cut longitudinally in four sections to maximize exposure and optimize histological view of the tubal plicae. The fallopian tube was cross-sectioned at intervals of 2–3 mm.19

All histological sections were stained with hematoxylin and eosin and reviewed by two gyneco-pathologists. In case of discrepancies, consensus was reached between pathologists. The columnar epithelium of the cervix was examined for atypia, presence of adenocarcinoma in situ, or carcinoma. The nature of the endometrium was classified in the following categories: atrophic endometrium, proliferative, secretory, disordered proliferative, hyperplasia, atypical hyperplasia, endometrial intraepithelial carcinoma, or invasive carcinoma.

The classification system for endometrial hyperplasia outlined by the World Health Organization (WHO) in 2003 was used in the current study.17, 18 Hyperplasia was defined as proliferating endometrium with architectural abnormalities such as budding and branching. Cells are mostly enlarged and exhibit nuclear pseudostratification, with an increased number of mitotic figures. An increased gland-to-stroma ratio (3:1) is required for diagnosis of hyperplasia, and it can present with or without atypical cells.20 In case of proliferative features, in absence of increased gland-to-stroma ratio, the endometrium was diagnosed as proliferative. The endometrium of postmenopausal women with proliferative features was considered disordered proliferative. Endometrial intraepithelial carcinoma was proposed by Ambros et al,21 and described as noninvasive, serous carcinoma resembling cells, that replace the endometrial surface epithelium and glands. An example of atypical hyperplasia and endometrial intraepithelial carcinoma is shown in Figure 2.

Figure 2
figure 2

(a) An example of serous carcinoma located in an ovary (level of magnification: × 100). The following pathologic lesions were identified in this cohort of women with serous ovarian carcinoma (level of magnification: × 100). (b) Serous tubal intraepithelial carcinoma located in the tubal epithelium, with adjacent benign tubal epithelium. (c) Endometrial intraepithelial carcinoma. (d) Atypical hyperplasia located in endometrial glands.

Full size image

The tubal sections were classified as follows: normal tubal epithelium, hyperplasia, minor epithelial atypia, serous tubal intraepithelial carcinoma, or invasive carcinoma.22 Tubal hyperplasia was defined as cellular crowding in absence of atypical nuclei, and minor epithelial atypia as cellular crowding, with slight cellular atypia and nuclei with small nucleoli, but without loss of polarity. Serous tubal intraepithelial carcinoma is defined as noninvasive, serous carcinoma resembling cells that replace the tubal epithelium.23, 24, 25 They are characterized by a high nuclear-to-cytoplasm ratio, nuclear pleomorphism, stratification, and a high mitotic index. Further, the presence of serous tubal intraepithelial carcinoma in the fimbrial part of the fallopian tube, the non-fimbrial part, or in both locations was recorded.

Histological Examination of the Ovaries

The ovaries were embedded following standard protocol and sections were scored for histological carcinoma type, the location and extent of the tumor, and the tumor differentiation grade (low or high grade).26, 27 Furthermore, concurrent borderline tumors in the ovaries or other abnormalities were scored.

Immunohistochemistry

Immunohistochemical staining was performed for all cases identified with an endometrial intraepithelial carcinoma, and coexisting serous tubal intraepithelial carcinoma and/or atypical hyperplasia. Representative sections of 4 μm were sliced from formalin-fixed paraffin-embedded samples and incubated with monoclonal antibodies against p53 (dilution 1:250; Clone DO-7; Neomarkers, Fremont, CA, USA), Ki-67 (dilution 1:100; Clone MIB-1; Dako, Glostrup, Denmark), and WT1 (dilution 1:20; Clone WT49, NCL-L-WT1-562, Leica Biosystems, Newcastle Upon Tyne, UK). Antigen retrieval was processed according to the manufacturers’ guidelines. Immunohistochemical stainings were interpreted by two of the authors (JB, AAGvT). Staining was scored as the percentage of positive nuclear tumor cells and classified as follows: 0%, <5%, 5–25%, 26–50%, 51–75%, >75%. Intensity of staining was scored as weak, moderate or strong.

Statistical Analysis

Statistical analyses were performed using SPSS software version 20.0 (SPSS, Chicago, IL, USA) and P<0.05 (two-sided test) was considered statistically significant. For comparison of groups of women with or without serous tubal and/or endometrial intraepithelial carcinoma, the Pearson’s chi-square (χ2) test was used, or when appropriate, the Fisher’s exact test. For comparison of continuous variables between both groups, the Mann-Whitney U test was used.

Results

Clinical Characteristics

The study cohort included 60 women diagnosed with high-grade serous ovarian cancer. Clinical characteristics of these women are shown in Table 1. Mean age was 63 years, the majority were postmenopausal (88%) and mean body mass index was 24 kg/m2. The FIGO stage distribution was as follows: 3 (5%) women had a FIGO stage I, 4 (7%) a stage II, 47 (80%) a stage III, and 5 (8%) a stage IV. Seventy-five percent of patients received chemotherapy before debulking surgery.

Table 1 Clinical characteristics of the serous ovarian cancer patients (N=60)
Full size table

Eleven (18%) patients had another cancer diagnosed before their diagnosis of ovarian carcinoma: eight were diagnosed with breast carcinoma, two with colon carcinoma and one with melanoma. In four cases, the breast carcinoma was diagnosed within 6 months from the ovarian carcinoma, and they were considered double primary carcinomas. In four (7%) cases of the study population, a BRCA1 mutation was identified, but analysis was not performed for all included women. Three BRCA1 mutation carriers also had a diagnosis of breast carcinoma.

Overview of Tissues Derived from the Müllerian duct in Patients with Serous Ovarian Cancer

In three cases, the cervix was not available for histological examination, because a supravaginal hysterectomy was performed at debulking surgery. No pathology could be found in the endocervix of women with serous ovarian carcinoma, except for one case with serous carcinoma expanding toward the cervix, and another case with severe dysplasia (CIN 3).

In Table 2, results are shown of the histological examination of the endometrial sections. The endometrium was diagnosed as benign endometrium in 38 (63%) cases: in 29 (48%) cases the endometrium was atrophic, (disordered) proliferative or secretory, and in 9 cases (15%) hyperplastic. In 13 (22%) cases the endometrium showed atypical hyperplasia, in 3 (5%) cases an endometrial intraepithelial carcinoma was identified, and in 6 (10%) endometrial intraepithelial carcinoma and atypical hyperplasia. In total, endometrial intraepithelial carcinoma was found in 9 (15%) cases and atypical hyperplasia in 19 cases (32%).

Table 2 Results of histological assessment of the endometrium in a cohort of serous ovarian cancer patients, sampled in accordance with the SEE-End (Sectioning and Extensively Examining of the Endometrium) protocol
Full size table

Concurrent invasive serous endometrial carcinoma was found in five cases, and all occurred in cases diagnosed with additionally endometrial intraepithelial carcinoma and/or atypical hyperplasia. Invasive endometrial carcinoma was present in 23% (5/22) of cases diagnosed with endometrial intraepithelial carcinoma and/or atypical hyperplasia. All invasive endometrial carcinomas were of serous type. In one case it was seen concurrent to atypical hyperplasia, in one case concurrent to endometrial intraepithelial carcinoma, and in three cases concurrent to both endometrial intraepithelial carcinoma and atypical hyperplasia. In one of the latter, the endometrial and ovarian carcinoma were interpreted as double primaries, with the largest primary carcinoma located in the endometrium and a second focus of carcinoma located adjacent to borderline carcinoma in the left ovary. Four cases were classified as primary ovarian carcinoma with metastasis in the endometrium, as the largest carcinoma mass was located in the ovary and diffuse ovarian involvement was seen.

Results of histological assessment of the tubal epithelium was shown in Table 3. In 18 (33%) cases the tubal epithelium was normal, in 2 (4%) cases hyperplasia was found, in 11 (20%) cases minor epithelial atypia, and in 23 (43%) cases serous tubal intraepithelial carcinoma was identified. Serous tubal intraepithelial carcinoma was identified in the fimbrial region of the fallopian tubes in 15 (65%) cases, in the non-fimbrial region in 4 (17%) cases, and in both fimbrial and non-fimbrial regions in 4 (17%) cases. In six (8%) cases the tubal epithelium was not available for histological assessment, in four cases the tubes could not be identified in the tumor mass, and in two cases the tubes were already removed during incomplete debulking surgery in another hospital.

Table 3 Results of the histological assessment of the fallopian tubal epithelium in a cohort of serous ovarian cancer patients, sampled in accordance with the SEE-Fim (Sectioning and Extensively Examining of the Fimbriated end) protocol
Full size table

In 50% (30/60) of cases with serous ovarian carcinoma, also invasive carcinoma was found in the tubal epithelium. The carcinoma was restricted to the fimbrial part of the tube in 28% (17/60) of cases. In 60% (18/30) of cases with serous invasive tubal carcinoma, additionally a serous intraepithelial carcinoma was found in the tubal epithelium, whereas serous tubal intraepithelial carcinoma was found in only 17% (5/30) of cases without serous tubal carcinoma.

Characteristics of Serous Ovarian Carcinoma with Endometrial Intraepithelial Carcinoma and/or Serous Tubal Intraepithelial Carcinoma

Clinical and pathological characteristics were compared between groups of women diagnosed with endometrial intraepithelial carcinoma and/or serous tubal intraepithelial carcinoma, and women diagnosed without these intraepithelial carcinomas (Table 4). None of the characteristics were significantly different between the two groups of women, except that the diagnosis of cancer before the ovarian carcinoma was more common in cases diagnosed with serous tubal and/or endometrial intraepithelial carcinomas (P=0.004). Cases who underwent chemotherapy before debulking surgery did not show significantly more often serous tubal and/or endometrial intraepithelial carcinoma (P=0.872). Further, bilateral ovarian carcinoma involvement, enlarged ovaries, presence of additional ovarian lesions as cysts, or type of ovarian carcinoma pattern (diffuse/multinodular/mainly surface) was not significantly different for cases diagnosed with or without serous tubal intraepithelial carcinoma and/or endometrial intraepithelial carcinoma.

Table 4 Comparison of characteristics between serous ovarian cancer patients with or without endometrial intraepithelial carcinoma (EIC) and/or serous tubal intraepithelial carcinoma (STIC) (N=54)
Full size table

Coexisting Endometrial Intraepithelial Carcinoma and Serous Tubal Intraepithelial Carcinoma in Müllerian-Derived Tissues of Serous Ovarian Cancer Patients

In Table 5, an overview is given of the histologically examined tissue of all cases in which a serous tubal intraepithelial carcinoma, endometrial intraepithelial carcinoma, and/or atypical hyperplasia was found concurrent to serous ovarian carcinoma. In 31 (52%) cases one or more of these lesions were identified in the analyzed tissues, in 9 (15%) cases serous tubal intraepithelial carcinoma was found, in 5 (8%) atypical hyperplasia, in 1 (2%) endometrial intraepithelial carcinoma, and in 16 (27%) cases multiple pathological lesions were found. In eight (13%) cases serous tubal intraepithelial carcinoma and atypical hyperplasia, in five (8%) cases endometrial intraepithelial carcinoma, as well as atypical hyperplasia and serous tubal intraepithelial carcinoma, in two (3%) cases endometrial intraepithelial carcinoma and atypical hyperplasia, and in one (2%) case serous tubal and endometrial intraepithelial carcinoma (Table 5).

Table 5 Overview of Müllerian-derived tissues in serous ovarian cancer patients in which a serous tubal intraepithelial carcinoma (STIC), endometrial intraepithelial carcinoma (EIC), and/or endometrial atypical hyperplasia was found
Full size table

Immunophenotypes of endometrial intraepithelial carcinoma were compared with phenotypes of coexisting serous tubal intraepithelial carcinoma and/or atypical hyperplasia (Table 6). Six cases analyzed with serous tubal intraepithelial carcinoma and endometrial intraepithelial carcinoma, showed almost comparable p53 as well as Ki-67 expression in coexisting lesions. WT1 showed a difference in expression of more than one expression class in three cases, and was comparable in two. In one of these cases, tissue of the serous tubal intraepithelial carcinoma was not available for additional incubation with monoclonal antibody against WT1. Immunophenotypes of coexisting endometrial intraepithelial carcinoma and atypical hyperplasia showed lower expression for p53 and Ki-67 in atypical hyperplasia compared with endometrial intraepithelial carcinoma, and WT1 was negative in all atypical hyperplasia, regardless of the WT1 expression in the endometrial intraepithelial carcinoma.

Table 6 Immunohistochemical expression patterns of endometrial intraepithelial carcinoma (EIC) and coexisting atypical hyperplasia (AH) and/or serous tubal intraepithelial carcinoma (STIC) in serous ovarian cancer patients
Full size table

Serous Tubal Intraepithelial Carcinoma, Endometrial Intraepithelial Carcinoma and/or Atypical Hyperplasia Occur Significantly more often Synchronously

In Table 7, the prevalence of serous tubal intraepithelial carcinoma is analyzed in concurrence to the different entities identified in the endometrium. Serous tubal intraepithelial carcinoma was significantly more often found in concurrence to endometrial intraepithelial carcinoma, atypical hyperplasia, and/or invasive endometrial carcinoma compared with normal endometrium (P=0.010; Pearson chi-square). In 64% of cases identified with serous tubal intraepithelial carcinoma, an endometrial intraepithelial carcinoma and/or atypical hyperplasia was found in the endometrium, whereas in 28% of cases with serous tubal intraepithelial carcinoma the endometrium was benign. Of the latter, in four cases non-atypical hyperplasia was found in the endometrium.

Table 7 The coexistence of serous tubal intraepithelial carcinoma (STIC) and endometrial intraepithelial carcinoma (EIC) and/or endometrial atypical hyperplasia in serous ovarian cancer patients (N=54)
Full size table

Discussion

High-grade serous ovarian carcinoma is suggested to originate from tissues embryologically derived from the Müllerian ducts, followed by spread into the abdominal cavity by exfoliation and migration of loose cohesive cells.23, 28 The fallopian tubes have been extensively examined in the last decade and serous tubal intraepithelial carcinoma has been found in serous ovarian carcinomas in a mean overall prevalence of 41%.6, 8, 9, 10 The current study is unique, as the entire epithelium of the endometrium and the tubal epithelium were extensively assessed, as well as representative sections of the cervix. In 52% of cases, a serous tubal intraepithelial carcinoma, endometrial intraepithelial carcinoma, and/or endometrial atypical hyperplasia was found. In the endocervical tissue, one CIN3 lesion was found. In the fallopian tubes, serous tubal intraepithelial carcinoma was present in 43%, and in the endometrium an intraepithelial carcinoma was found in 15% of cases and atypical hyperplasia in 32%. In a quarter of cases, a lesion was found in the tubes as well as in the endometrium. Serous tubal intraepithelial carcinoma occurred significantly more often in women diagnosed with an endometrial intraepithelial carcinoma and/or atypical hyperplasia compared with women with a benign endometrium, regardless of the type of endometrial lesion (P=0.010).

To the best of our knowledge, the surprisingly high prevalence of atypical hyperplasia and endometrial intraepithelial carcinoma in serous ovarian cancer patients has not been previously described. The few studies that investigated the endometrium in pathology reports of asymptomatic women, reported a much lower prevalence of atypical hyperplasia (0.5–1.1%) and no endometrial intraepithelial carcinoma.29, 30, 31 We recently reported the prevalence of atypical hyperplasia in 6% of women who underwent a hysterectomy because of uterovaginal prolapse, after embedding the endometrial tissue in accordance with the SEE-End protocol.32 In none of these women, endometrial intraepithelial carcinoma was found.

The prevalence of serous tubal intraepithelial carcinoma in the current study was comparable to other studies that extensively examined the fallopian tubes in accordance with the SEE-Fim protocol.6, 9 In these studies, serous tubal intraepithelial carcinoma, was found restrictive to ovarian cancers of serous histology type, which supports its suggested role in the development of serous carcinoma.10, 33 Tubal hyperplasia and minor epithelial atypia were equally present in women at risk of ovarian carcinoma compared with women from normal populations, and considered as variants of normal tubal epithelial proliferation.22

Recently, the introduction of the SEE-Fim protocol resulted in a significant increase of the identification of serous tubal intraepithelial carcinoma, as it enabled optimal microscopic view of the tubal plicae.19 The endometrium is not particularly difficult to visualize microscopically, but its heterogeneous aspect and the small size of some lesions necessitates meticulous screening.20 The introduction of the SEE-End protocol might contribute to a more accurate identification of endometrial pathology, as shown by the higher prevalence of endometrial pathology found in asymptomatic women after introduction of the new protocol.32 Especially for the identification of endometrial intraepithelial carcinoma, the SEE-End protocol is recommended, as it can occur only unifocal and can be easily missed with routine sampling.

Further, identification of invasive endometrial carcinoma seems slightly increased in the current study (8%), compared with the prevalence reported after routine sampling (2–5%).15, 34 This might be because of the introduction of the SEE-End protocol. In half of the cases, additional serous tubal carcinoma was found and in one case was found even in the endocervix. Prevalence of extensive carcinoma is high and we cannot exclude that some of these cases might be misclassified as primary ovarian instead of primary tubal or endometrial. However, diagnostic criteria for extensive serous carcinoma have not been well defined.

The high prevalence of endometrial atypical hyperplasia in this group of serous ovarian cancer patients is remarkable, as atypical hyperplasia is known as a precursor of endometrioid endometrial carcinoma and its development is influenced by genetic predisposition and increased, unopposed, estrogen.35 Some endometrioid ovarian cancers have been described to cause increased estrogen levels, and this has even been reported for a few cases with serous ovarian cancer.36, 37, 38 However, we have no information on estrogen levels from cases included in the current study.

Endometrial intraepithelial carcinoma is involved in the development of serous endometrial carcinoma, and is occasionally found concurrent to serous ovarian or extra-ovarian carcinoma, without presence of invasive endometrial carcinoma.11, 39, 40, 41 An endometrial intraepithelial carcinoma has the capability of disseminating throughout the peritoneal cavity, and could putatively be responsible for a proportion of serous epithelial ovarian cancers.42, 43

Interestingly, we often identified simultaneous prevalence of serous tubal intraepithelial carcinoma, endometrial intraepithelial carcinoma, and/or atypical hyperplasia in patients with serous ovarian carcinoma. Several hypotheses can be put forward to explain their coexistence. Endometrial atypical hyperplasia represents most likely a synchronous precursor of endometrioid endometrial carcinoma in these women with serous ovarian carcinoma, instead of a lesion involved in the development of serous ovarian cancer. This is supported by the different immunophenotypes for atypical hyperplasia and coexisting endometrial intraepithelial carcinoma. However, serous tubal intraepithelial carcinoma and endometrial intraepithelial carcinoma are both involved in the development of serous carcinoma, and their coexistence is in contrast with the assumption that serous carcinoma develops from one location, from where it can develop into invasive carcinoma or spread to other locations in the peritoneal cavity.

An explanation for coexisting serous tubal and endometrial intraepithelial carcinomas is that one lesion represents the primary location from where the tumor spreads, and that the other lesion is a direct product of implantation of migrating cells. Spread of tumor cells from an endometrial intraepithelial carcinoma into the uterine cavity, through the tubes, toward the peritoneal cavity seems a plausible explanation, as it follows the route of retrograde menstruation.11, 43, 44 However, as suggested by Jarboe et al,40 serous tubal intraepithelial carcinoma might also represent the primary location, as cells may exfoliate toward the peritoneal cavity as well as toward the uterine cavity.45 They reported on a few cases diagnosed as endometrial serous carcinoma with coexistence of serous tubal intraepithelial carcinoma and endometrial intraepithelial carcinoma.40 However, most cases were without invasive endometrial carcinoma, but with tumor involvement of both ovaries. Therefore, cases seem comparable to the ones diagnosed in the current study as serous ovarian cancer with coexisting serous tubal and endometrial intraepithelial carcinomas. Jarboe et al,40 analyzed the tumor origin of four cases with coexisting intraepithelial carcinomas with immunohistochemical Wilms tumor 1 staining. Results were inconclusive and comparable to ours, as two cases showed strong expression in both lesions, whereas one case showed heterogeneous expression and another one none.40

Another intriguing explanation for the coexistence of serous tubal intraepithelial carcinoma and endometrial intraepithelial carcinoma is that carcinoma development does not occur in only one specific area of the Müllerian duct, but that the entire tract is susceptible for tumor development in these women. This could explain for the discordant WT1 staining in some of the coexisting serous tubal and endometrial intraepithelial carcinomas. A similar mechanism of field carcinogenesis has been described in many cancers, as for example in colonic and head and neck cancers, and has occasionally been described for ovarian cancers.46, 47, 48 This mechanism could also explain that gynecologic carcinomas, of comparable or different histological subtypes, are often identified synchronously. Additional molecular genetic studies are necessary to differentiate primary carcinomas from metastases and to elucidate the route of carcinogenesis in these women.

Clinical and pathological characteristics were almost comparable for women with and without a serous tubal and/or endometrial intraepithelial carcinoma in the current study. However, clinical variables were retrieved retrospectively and not always complete. The majority of women included in this study underwent chemotherapy before debulking surgery. Although, no significant difference in prevalence of endometrial intraepithelial carcinoma, serous tubal intraepithelial carcinoma or atypical hyperplasia was found between both groups, we cannot exclude that chemotherapy might have an impact on the prevalence of pathologic lesions. Another limitation of the current study is that the cervical tissue was not completely embedded.

In conclusion, the current study provided an overview of pathology in tissues derived from the Müllerian duct in women with serous ovarian cancer, embedded in accordance with the SEE-Fim and SEE-End protocol. Atypical hyperplasia, serous tubal intraepithelial carcinoma, and endometrial intraepithelial carcinoma were commonly identified in these women, and often occurred simultaneously. The clinical significance and biological potential of these lesions in women diagnosed with serous ovarian carcinoma may be of utmost importance in understanding tumor development of ovarian cancers.