Incidence/mortality

Uterine corpus cancer is the third most common cancer in women in the United States, with approximately 65,620 new cases in the United States in 2020 and 12,590 estimated deaths. Endometrial cancers can broadly be classified into two groups. Cancers of endometrioid histology are common (about 80% of cases), are associated with obesity and estrogen exposure, and generally have good prognosis. Nonendometrioid cancers are more rare (~ 20%), are less hormone dependent, and have poorer prognosis.

Endometrial cancer incidence has been steadily rising in the United States and many other areas internationally, with an average annual percent change in the United States of 1.4%–3.1%. Analysis of the SEER database in the United States has suggested that this increase is largely driven by nonendometrioid cancers.

Uterine serous carcinoma is the most common form of nonendometrioid uterine corpus cancer. Data from the Surveillance, Epidemiology, and End Results (SEER) database from 1988 to 2001 suggest uterine serous carcinoma accounted for about 10% of uterine cancers but 39% of deaths. Uterine serous carcinoma is more likely to be stage III or IV at presentation (52% of cases), even compared with other high-risk endometrial cancer histologies (36% for clear cell and 29% for grade 3 endometrioid carcinoma). Median age at diagnosis is in the 60s ( Table 12.1 ).

Etiology/risk factors

When first described, uterine serous carcinoma was noted for its histological similarity to ovarian papillary serous carcinoma. Prominent features included marked nuclear pleomorphism, complex papillary architecture, fibrous stalks, prominent tumor necrosis, and frequent psammoma bodies. Additionally, disease was often noted within the myometrium, ovarian lymphatics, or vascular channels without evidence of gross disease.

Later studies also observed that within uterine serous carcinoma, different histological subtypes are apparent, including disease admixed with other histologies such as endometrioid or clear cell and disease confined to endometrial polyps. Overall, the authors found that these tumors were clinically similar in their aggressive behavior and shared hobnail-shaped cells with high-grade nuclei with foci of papillary or glandular architecture. These findings led them to propose renaming the disease to uterine serous carcinoma. Subsequent studies have found that mixed adenocarcinomas with serous components, even if they contained some endometrioid or clear cell components, should be treated similarly to uterine serous carcinoma as the presence of a serous component was associated with worse prognosis.

The relationship between myometrial invasion and disease severity also differs in uterine serous carcinoma compared to endometrioid endometrial cancer. While extrauterine spread is rare in low-grade endometrioid endometrial cancer with limited myometrial invasion and small tumor size, early studies found that patients with non-invasive uterine serous carcinoma had a high risk of stages III–IV disease and high rates of death from disease. Subsequent studies have affirmed these findings. Noninvasive or early invasive forms of uterine serous carcinoma have been variably termed “intraepithelial serous carcinoma” and “endometrial intraepithelial carcinoma” or “superficial serous carcinoma,” respectively, and have previously been felt to represent earlier forms of disease. However, in one study, 3 of 9 (33%) patients with noninvasive disease and 15 of 31 (48%) patients with early invasive disease had extrauterine spread. Considering these groups together as “minimal uterine serous carcinoma” to highlight the invasive potential of these uterine lesions, 5-year survival was 94% for patients without extrauterine spread vs 56% with extrauterine spread. Thus, even for apparently polyp-confined disease, complete surgical staging with detailed pathological scrutiny to rule out extrauterine spread is advised. Though noninvasive disease does not rule out extrauterine spread, larger studies have found that depth of myometrial invasion does predict worse outcome, with 38% of patients with noninvasive disease having stages III–IV disease vs 80% of patients with depth of invasion ≥ 50%.

Efforts at identifying a true precursor lesion to uterine serous carcinoma led to description of a distinct entity termed endometrial glandular dysplasia in 2004. This is morphologically distinct from both benign endometrium and uterine serous carcinoma, has molecular features shared with subsequently diagnosed uterine serous carcinoma, and can be diagnosed by EMB or D&C. The clinical implications of this pathological entity are still unclear, but point to key steps in the development of uterine serous carcinoma.

New capabilities in molecular characterization have further helped elucidate the biological underpinnings of uterine serous carcinoma. Mutations in TP53 are the most common genetic alteration in uterine serous carcinoma, present in over 90% of cases. While TP53 mutations are never observed in resting endometrium, they occur in 43% of endometrial glandular dysplasia and 72% of serous endometrial intraepithelial carcinoma, suggesting an early role of TP53 in uterine serous carcinoma carcinogenesis. An even earlier lesion termed “p53 signature glands” may represent an even earlier form of disease identifiable only by immunoassay rather than histological change. As TP53 is the most commonly mutated gene in human cancer, improved understanding of the biology of uterine serous carcinoma may yield insights into other cancer types, and vice versa. Other recurrently mutated genes in uterine serous carcinoma include PIK3CA , FBXW7 , and PPP2R1A.

Risk factors for uterine serous carcinoma include African-American race, older age, and multiparity. Overall, uterine serous carcinoma is approximately twice as common in non-Hispanic black women than in other populations (non-Hispanic white, Hispanic, Asian) in a US study population from 2000 to 2011. In addition, while rates of low-grade endometrioid endometrial cancer decreased slightly in non-Hispanic whites (annual percent change or APC = − 0.8) and increased slightly in non-Hispanic blacks (APC = 1.0), rates of uterine serous carcinoma are increasing in all racial groups. However, the increase in rate is more pronounced in minority populations, with rates of increase ranging from APC = 2.8 in non-Hispanic Whites to 3.8 in non-Hispanic blacks to 4.5 in Hispanics and 9.0 in Asians. Non-Hispanic blacks face worse outcomes in endometrial cancer even controlling for histology and stage of disease, and this trend sadly is also true in uterine serous carcinoma. Non-Hispanic white, Asian, and Hispanic populations experience 5-year survival rates of about 80%, 40%, and 20% for localized, regional, and distant disease. Analogous figures for non-Hispanic black populations are approximately 70%, 30%, and 10%. The reasons behind the rising rates and racial disparities seen in uterine serous carcinoma are poorly understood and likely multifactorial.

An area of much controversy is the association between BRCA1/2 alterations and risk of development of uterine serous carcinoma. Past studies have linked a history of breast cancer with increased risk of subsequent development of uterine serous carcinoma. BRCA1 alterations are disproportionately common in patients with uterine serous carcinoma and may predispose to both breast and uterine cancers. Specifically, loss-of-heterozygosity events have been seen in uterine serous carcinoma in a BRCA1 germline carrier, suggesting a causal role for this mutation. That said, the risk of developing uterine serous carcinoma in women who are germline positive for BRCA remains undefined. The recommendation for a hysterectomy at the time of risk reducing bilateral salpingo-oophorectomy (RR-BSO) in women with germline BRCA1 mutation remains controversial but should be discussed with patients.

The oncogenic potential of the human epidermal growth factor receptor 2 (HER2) is well known. HER2, encoded by the gene ERBB2 , is receptor tyrosine kinase. Among all four HER family proteins, HER2 has the strongest catalytic kinase activity and functions as the most active signaling complex of the HER family after dimerization with other HER family members [1, 2]. Overexpression of HER2 leads to increased homodimerization (HER2:HER2) and heterodimerization (e.g., HER2:HER3), which initiates a strong pro-tumorigenic signaling cascade [3]. Overexpression of this oncogene plays an important role in the development and progression of certain aggressive types of breast, gastric and uterine cancers [4–7]. While the protein has become an important biomarker and target of therapy for approximately 30% of breast cancer patients, has also been identified that 25%–30% of women with uterine serous carcinoma also overexpress HER2.

Gross appearance

Serous carcinoma of the endometrium most commonly occurs in atrophic uteri and can present with grossly identifiable tumor or as microscopic disease, often involving endometrial polyps. Myometrial invasion as well as adnexal involvement may be grossly evident or only identified microscopically.

Microscopic features

Uterine serous carcinoma can demonstrate glandular ( Fig. 12.1 A), papillary ( Fig. 12.1 B) or solid architecture. Although there can be substantial variability in the cytologic features, marked cytologic atypia is a hallmark of serous carcinoma ( Fig. 12.1 A ). The tumor cells may have eosinophilic cytoplasm or cytoplasmic clearing, sometimes with hobnail features. Many, if not all, of the tumor cells show nuclear pleomorphism, hyperchromasia, and prominent eosinophilic nucleoli. Mitotic figures, including abnormal mitotic figures, are easily identified, as are smudged nuclei. When serous carcinoma arises in an endometrial polyp, the tumor cells line the surface of the polyp and may show extension into underlying glands with no associated stromal invasion. This phenomenon can also be seen in the native endometrium. This lesion was previously called serous endometrial intraepithelial carcinoma. However, given that the purely intraepithelial carcinoma can be associated with metastatic disease, it is now recommended to simply refer to it as serous carcinoma. Uterine serous carcinomas are all considered high grade, thus the use of “high grade” to further describe these tumors is considered redundant and not recommended. Lymphovascular invasion can be focal or extensive. Careful attention to exclude other mixed components including endometrioid carcinoma, clear cell carcinoma and sarcoma is necessary.

(A) Glandular pattern of uterine serous carcinoma (USC) with marked cytologic atypia. (B) Papillary pattern of USC with marked nuclear heterochromasia. (C ) Immunohistochemistry for p53 showing an aberrant (strong, diffuse) staining pattern. (D) Immunohistochemistry for p53 showing an aberrant (null) staining pattern.

Immuohistochemistry

Serous carcinomas express epithelial markers such as keratins and EMA, PAX-8, and hormone receptors though the latter show decreased expression compared to endometrioid carcinomas. The most common pattern of aberrant expression of p53 is diffuse strong nuclear staining in > 75% of the tumor cells ( Fig. 12.1 C). However, in a subset of cases, there is complete loss of staining consistent with a null-phenotype ( Fig. 12.1 D). In the latter, positive staining in internal controls is necessary to ensure authenticity of the absent staining. More recently diffuse cytoplasmic staining has also been recognized as a pattern of aberrant p53 expression.

Differential diagnosis

The most critical differential is distinguishing serous carcinoma from any low-grade form of endometrioid carcinoma which can also demonstrate papillary and glandular growth patterns. However, the presence of marked cytologic atypia and frequent mitoses favor serous carcinoma. Ancillary p53 and p16 immunohistochemical (IHC) studies can help in this distinction. The vast majority of endometrial serous carcinomas show aberrant staining for p53 along with intense, block-like staining for p16, whereas the vast majority of low-grade endometrioid tumors will show wild-type p53 staining and heterogeneous staining for p16. It is important to record the IHC stains in a manner that can be definitively interpreted from the report. It is recommended that p53 IHC be reported as aberrant with either diffuse, strong nuclear staining or as a complete lack (null-pattern) of staining. Both staining patterns are reliable predictors of TP53 mutations in serous carcinoma. Results of p16 IHC should be reported as diffuse, block-like staining or as patchy staining. In addition, some serous carcinomas show overlapping features with clear cell carcinoma, as serous carcinoma can show cytoplasmic clearing and clear cell carcinoma can consist largely of cells with eosinophilic cytoplasm similar to serous carcinoma. However, serous carcinomas lack a tubulocystic growth pattern and eosinophilic, hyalinized stroma usually seen in clear cell carcinoma. In addition, immunohistochemistry can aid in this distinction as clear cell carcinomas are usually negative for ER/PR expression while many serous carcinomas are at least focally positive. In addition, HNF-1B and napsin-A are often expressed in clear cell carcinoma. Of note, p53 staining can show aberrant expression in a subset of clear cell carcinomas.

Molecular findings

Serous carcinomas are defined by alterations in the TP53 gene. The high incidence of this alteration results in the usefulness of p53 immunohistochemistry as an ancillary test in the diagnosis of serous carcinoma. In over 75% of cases p53 immunohistochemistry will show an aberrant staining pattern usually with strong, diffuse nuclear staining with fewer cases showing a complete absence (null pattern) of staining. In addition, the genomes of serous carcinomas exhibit numerous copy number alterations, hence the designation as copy number high tumors in The Cancer Genome Atlas (TCGA) classification of endometrial carcinoma. A number of other genes ( PIK3CA , PPP2R1A , FBXW7 , PIK3R1 ) are commonly mutated in serous carcinoma and CCNE1 is amplified in approximately 20% of cases. Importantly, approximately 18% of serous carcinomas show amplification of ERBB2 and are, thus, potentially amenable to targeted therapy. Consequently, HER2 immunostaining with HER2 FISH studies, if indicated, are recommended for all metastatic or recurrent cases of serous carcinoma.

Differential diagnosis

Uterine serous carcinoma presents similarly to other uterine adenocarcinomas and is indistinguishable based on signs and symptoms alone; however, an accurate diagnosis is essential as the treatment and prognosis in endometrial adenocarcinoma vs uterine serous carcinoma is significantly different. The diagnosis of uterine serous carcinoma is ultimately a histologic diagnosis (see Section “ Pathology ” above). The differential diagnosis for postmenopausal bleeding is vast and ranges from benign etiologies such as endometrial hyperplasia, polyps, and atrophy to malignant etiologies, including uterine serous carcinoma.

Signs and symptoms

Women with uterine serous carcinoma typically have disease detected after presentation with postmenopausal bleeding, abdominal pain, or uterine enlargement. Uterine serous carcinoma is an aggressive, fast-growing cancer, and as such, many patients with uterine serous carcinoma have extrauterine disease at presentation. Women with metastatic spread may develop pain or weight loss in addition to the above-mentioned symptoms.

Physical exam findings

As with other forms of endometrial cancer, postmenopausal bleeding is the most common presenting symptom for uterine serous carcinoma. Because many women with uterine serous carcinoma have extrauterine disease at diagnosis, presenting symptoms may also include abdominal pain or bloating as in ovarian serous carcinomas. Differential diagnosis includes other uterine/endometrial cancers, including carcinosarcoma, which often include a serous epithelial component.

Tumor markers

Given the histological similarities between epithelial ovarian cancer and uterine serous carcinoma, the serum marker cancer antigen 125 (CA-125) has been proposed as a marker of uterine serous carcinoma disease activity, though this is controversial. Three retrospective studies of CA-125 in the preoperative, posttreatment, and recurrent setting have found that as in epithelial ovarian cancer, pretreatment CA-125 levels are positively correlated with risk of extrauterine and recurrent disease, CA-125 falls substantially following initial treatment, and CA-125 rise after treatment is correlated with disease recurrence. However, there are no prospective studies evaluating CA-125 monitoring in uterine serous carcinoma, and whether CA-125 monitoring improves overall or progression-free survival (PFS) in this disease is unclear. However, our practice is to monitor CA-125 as part of surveillance after initial treatment.

Imaging

Given the high risk of extrauterine spread even in apparently polyp-confined disease, imaging with CT of the chest, abdomen, and pelvis or whole-body positron emission tomography (PET) can be considered to evaluate for the presence of metastatic disease. Magnetic resonance imaging (MRI) of the pelvis may also be considered. Unlike in endometrioid cancer, where the primary role goal of preoperative MRI is to assess risk of lymphatic spread by measuring depth of myometrial invasion, endometrial cavity thickness and tumor thickness, and presence of lymphadenopathy, the goal of MRI in uterine serous carcinoma is to assess for signs of invasion into the cervical stroma, parametrial structures, or bladder and rectum that may impact initial surgical planning. In the initial work-up of a newly diagnosed patient, we typically perform CT scan of the chest, abdomen, and pelvis. We may add a MRI if there is concern for cervical or parametrial involvement on physical exam or CT scan ( Table 12.2 ).

Diagnostic tests

Office-based Pipelle biopsy has approximately 99% sensitivity for detection of uterine serous carcinoma, though specificity is lower due to potential under-sampling disease that features mixed histology with clear cell or endometrioid components as described above. Additionally, ultrasound may not be able to exclude uterine serous carcinoma as a cause of postmenopausal bleeding even when endometrial stripe thickness is < 5 mm, as has been described for endometrioid endometrial cancers.