Endometrial Polyp in Postmenopausal Women: An Epicenter for the Development of Endometrial Serous Carcinoma Open Access

Serous endometrial intraepithelial carcinoma (SEIC), formerly endometrial intraepithelial carcinoma (EIC), is the earliest recognizable form of the disease.¹²  SEIC is defined as replacement of endometrial surface or underlying glandular epithelium by flat or micropapillary proliferation of high-grade malignant epithelial cells that have cytologic and immunohistochemical profiles of the tumor cells seen in a fully developed serous carcinoma (Figure 1, A and B). Although technically it is an in situ/intraepithelial malignancy without stromal invasion, SEIC carries a significant risk of high-stage disease at presentation.

Minimal uterine serous carcinoma (MUSC) is a diagnostic terminology that was proposed by Wheeler et al⁶  to provide an umbrella category of early endometrial serous carcinoma to include SEIC and superficial serous carcinoma (SSC). SSC is defined as an early invasive serous carcinoma with confluent glandular growth and/or stromal desmoplasia that is confined to the endometrium proper or an endometrial polyp with an overall tumor size of 1 cm or less (Figure 1, C and D). MUSC is a useful classification in defining these early forms of endometrial serous carcinoma for several reasons: first, it is often difficult to separate SEIC from SSC microscopically. Second, both lesions share a comparable rate of extrauterine disease at presentation. Third, SSC almost always coexists with SEIC, consistent with stepwise tumor progression. Investigations into these early forms of endometrial serous carcinoma are important to elucidate the pathogenesis of endometrial serous carcinoma.

Endometrial polyp is a common lesion in endometrial biopsy and hysterectomy specimens, and its incidence in postmenopausal patients is estimated to be 20%.¹³  Histologically, endometrial polyps in postmenopausal women are localized overgrowths of cystically dilated endometrial glands, fibrous stroma, and thick-walled vessels that form a polypoid or pedunculated endometrial surface mass. The prevalence of malignancy and premalignant lesions in endometrial polyps is 3% to 5.4% in the postmenopausal setting, and various histologic types of endometrial cancers can involve endometrial polyp, including serous, endometrioid, clear cell, and malignant mixed Müllerian tumor (carcinosarcoma).^14–16  Tamoxifen treatment for breast cancer incurs a 3% rate of carcinoma arising from an endometrial polyp,¹⁷  particularly endometrial serous carcinoma.¹⁸ 

The earliest recognition of a strong relationship between endometrial polyp and endometrial serous carcinoma was reported in 16 cases, of which 6 patients (37.5%) presented with concurrent extrauterine disease despite an absence of, or only superficial, myometrial invasion.¹⁹  In a study of 13 stage IA serous carcinomas, 6 tumors were solely identified within an endometrial polyp and an additional 4 tumors involved both an endometrial polyp and background endometrium.²⁰  Wheeler et al⁶  established the definition of MUSC in their report of 21 cases of the condition, of which 10 were confined to an endometrial polyp and an additional 9 involved both an endometrial polyp and background endometrium. McCluggage et al²¹  described 5 cases of uterine serous carcinoma involving an endometrial polyp, 2 of which were confined to the endometrial polyp. In another study of 13 cases of nonmyometrial invasive uterine serous carcinoma involving an endometrial polyp, all except one were limited to the polyp.¹²  Subsequently, our group reported 40 cases of MUSC, of which 35 tumors (88%) involved an endometrial polyp and 21 tumors (53%) were confined to an endometrial polyp.⁴  In another series of 8 cases of MUSC involving an endometrial polyp, 4 tumors were confined to the polyp.²²  In one larger investigation of 44 patients with stage IA endometrial serous carcinomas, including MUSC and those beyond the size limit of MUSC, the tumor was confined to an endometrial polyp in 17 cases (30.9%).²³  More recently, a multi-institutional series in Italy reviewed 75 serous carcinomas involving a polyp, including 56 polyp-confined cases (75%) and 19 cases with concurrent involvement of both the endometrial polyp and the background endometrial mucosa.²⁴  In the largest series of MUSC to date,¹¹  among 101 consecutive, fully staged pure serous carcinomas without myometrial invasion, we found that 77 MUSCs (76%; 77 of 101) involved an endometrial polyp, of which 68.8% (53 of 77) of the tumors were confined to the polyp. In summary, a significant number of studies indicate that as many as 90% of MUSCs involve an endometrial polyp, and 40% to 70% of MUSCs are confined to the polyp at the time of staging surgery. Such a close topographic relationship supports the notion that at least a subset of endometrial polyps in postmenopausal women represent an epicenter for the development of endometrial serous carcinoma.

p53 mutation is the fundamental oncogenic event essentially present in all endometrial serous carcinomas, including MUSCs. Immunohistochemistry for p53 is the most frequently used biomarker in the confirmation of an endometrial serous carcinoma. Abnormal p53 staining patterns are well correlated with p53 mutations, including diffuse and strong nuclear staining, complete loss of nuclear staining in all tumor cells (p53 “null”), and diffuse cytoplasmic staining with weak nuclear expression.

Endometrial polyp has been a subject of investigation for the pathogenesis of endometrial cancers.^25,26  Investigations into the histologic compartments adjacent to serous carcinoma have identified a few possible precursor lesions of endometrial serous carcinoma, including the p53 signature and endometrial glandular dysplasia.²⁷  The p53 signature is first described in the fallopian tube as a precursor lesion to tubal serous carcinoma.²⁸  Similar to the fallopian tube lesion, p53 signature in the endometrium consists of nonatypical epithelium that shows abnormal p53 staining patterns involving 12 or more contiguous epithelial cells (Figure 2, A and B) without a concurrent significant increase in the Ki-67 labeling index.^29,30  The p53 signature is frequently found adjacent to MUSC. One study found the presence of frequent p53 signature in endometrial polyps in postmenopausal women but not in endometrial polyps of premenopausal patients.²⁹  Endometrial glandular dysplasia is defined as endometrial glandular epithelial atypia that falls short of a diagnosis of SEIC^31,32  and is often present in endometrial polyps involved by MUSCs,^30,33  further supporting the theory that a subset of endometrial polyps may serve as a high-risk microenvironment where related precursor lesions may develop and transform into MUSC over time. Future studies targeting various histopathologic compartments of the endometrial polyp involved by MUSC and using comprehensive molecular genetic analysis, including next-generation sequencing, may elucidate the key steps of pathogenesis of endometrial serous carcinoma.

Although technically MUSC is a minimal malignancy, it carries a significant risk of high-stage disease at presentation. Extrauterine spread of serous carcinoma without myometrial invasion ranges from 45% to 62% at the initial surgery.^4–7  The extrauterine metastasis commonly involves surfaces of the ovary, fallopian tube, omentum, pelvic peritoneum, and, rarely, pelvic lymph nodes. Although it is possible that synchronous pelvic high-grade serous carcinoma arising from the ovary, fallopian tube, or peritoneum is an alternative explanation, the clonal, metastatic nature of the extrauterine tumors in patients with MUSC has been confirmed by several in-depth immunohistochemical and/or molecular mutational studies.^34–37  In our recent study, patients with MUSC involving both the endometrial polyp and background endometrium had a higher rate of extrauterine disease and lymphovascular invasion compared with patients with MUSC confined to the polyp.¹¹  It can be hypothesized that once the serous carcinoma cells from a polyp-confined MUSC extend to the background endometrium, the risk of extrauterine spread increases. Although shedding of tumor cells from MUSCs into the peritoneal cavity through the trans-tubal pathway is a favored mechanism for the extrauterine metastasis, the lymphovascular pathway remains as an alternative. The presence of lymphovascular involvement by MUSC significantly correlates with a higher tumor stage in our recent study.¹¹  Such significant correlation was also observed in several other investigations in the past,^38,39  although not in a few others.^8,23  Sample size and inclusion criteria may explain such conflicting findings. It is interesting to note that in one recent study, increased size of endometrial polyp associated with endometrial serous carcinomas including MUSC was correlated with a high-stage disease only in the presence of myometrial invasion.⁴⁰ 

It is relevant to note that several recent studies have emphasized the importance of serous carcinomas arising in an endometrial polyp as a unique and possibly more aggressive subtype of endometrial serous carcinoma, particularly in reference to Wilms tumor gene (WT-1) expression.^24,41–43  In one investigation, upon comparing 37 polyp-associated serous carcinomas and 25 non–polyp-associated counterparts, it was found that the former had a significantly higher rate of ovarian metastasis and a low rate of myometrial invasion, suggesting that serous carcinoma arising in an endometrial polyp may represent a distinct subtype of the disease.⁴¹  Polyp-associated serous carcinomas have more frequent WT-1 expression by immunohistochemistry than those without involvement of an endometrial polyp,⁴¹  and the expression of WT-1 in endometrial serous carcinoma predicated a worse disease-free survival in another study of 77 cases.⁴⁴ 

Given the close topographic relationship between MUSC and endometrial polyp and the fact that even a minute focus of MUSC can spread early to extrauterine sites, leading to a dismal clinical outcome, timely recognition, accurate diagnosis, and reporting of MUSC are of paramount importance.⁴⁵  It is important to realize that SEIC is a form of endometrial serous carcinoma, although technically it is an in situ or intraepithelial carcinoma.⁴⁶  The clinical management of the patient does not differ regardless of whether the biopsy diagnosis is SEIC, SSC, or full-blown serous carcinoma. For this reason, the author recommends using minimal uterine serous carcinoma (MUSC) as the diagnostic terminology to correctly convey the critical clinical implication and discourages the use of “in-situ” serous carcinoma or even SEIC to avoid potential misunderstanding and therefore undertreatment of the patient by clinical colleagues.

Diagnosis of serous carcinoma, particularly MUSC in small biopsy or limited curettage specimens, may be difficult.⁴⁷  Papillary proliferation, cell stratification, irregular/uneven luminal borders, with budding or forming slitlike glandular spaces, are characteristic features of endometrial serous carcinoma. A solid growth pattern can also be seen in some cases, although this is more often admixed with the typical histologic patterns. The tumor cells are polygonal, cuboidal, or columnar, with marked cytologic atypia including high nuclear to cytoplasmic ratio, nuclear pleomorphism, dark or coarse chromatin, and macronucleoli (Figure 1). Mitotic figures are brisk, with frequently abnormal forms. The presence of nuclear polarity along with a smooth luminal border with epical cytoplasmic clearing are features of endometrioid differentiation, although they can be seen focally in serous carcinoma in rare cases. Well-formed glandular lumen in the presence of grade 3 tumor cell nuclei should prompt a thorough effort to rule out serous carcinoma.⁴⁵  MUSC may also have focal cytoplasmic clearing and/or hobnailing, overlapping with clear cell carcinoma. The application and appropriate interpretation of immunomarkers, including p53, p16, estrogen receptor (ER), progesterone receptor (PR), phosphatase and tensin homolog, Napsin A, and hepatocyte nuclear factor-1β (HNF-1β) can be very helpful. Diagnosis of MUSC in a small biopsy or limited curettage specimen demands careful microscopic examination by pathologists, particularly in the presence of endometrial polyp in a postmenopausal patient. Although WT-1 expression is a characteristic marker for fallopian tube and ovarian serous carcinomas, a significant percentage (up to 71%) of MUSCs⁴⁸  and (27%) endometrial serous carcinomas in general⁴³  express this biomarker. A recent study found that polyp-associated serous carcinomas had an immunohistochemical profile (ER⁺, PR⁺, Napsin A⁺, and WT-1⁺) different from that of non–polyp-associated serous carcinomas (WT-1⁻ and Napsin A⁺ and low-level expression of ER and PR).⁴¹ 

A close topographic relationship between MUSC and endometrial polyp has been established by many studies, supporting the theory that most, if not all, early endometrial serous carcinomas arise in an endometrial polyp in postmenopausal women. MUSCs arising from an endometrial polyp may represent a distinct, more aggressive subtype of serous carcinoma. Future investigations into various compartments of endometrial polyp involved by MUSC may elucidate the precursor lesions and key molecular alterations underpinning the development of endometrial serous carcinoma. Given the high-risk behavior of MUSC, the importance of early pathologic diagnosis and timely comprehensive staging surgery cannot be overemphasized.