Clinical case
A 34-year-old female presents with worsening abdominal distension. She is found to have a 22 cm solid and cystic pelvic mass likely arising from the right ovary on MRI, and no evidence of metastatic disease ( Fig. 5.1 ). CA-125 is within normal range for a premenopausal female at 36 (ULN 46 U/mL), hCG and prolactin levels are normal, and CEA is elevated to 15 (ULN 5 ng/mL). The patient undergoes surgical removal of the mass, right tube and ovary via laparotomy. The mass is removed intact with no evidence of intraoperative rupture. On histopathological evaluation, the tumor is determined to be a primary mucinous ovarian carcinoma with expansile-type pattern of invasion. She desires fertility preservation.
How do you approach the management of this patient?
Epidemiology
Incidence/mortality
Mucinous carcinomas of the ovary are a rare subtype of epithelial ovarian carcinomas, accounting for less than 3% of all epithelial ovarian cancers ( Table 5.1 ). Historical studies cite the frequency of primary mucinous ovarian carcinoma ranging from 6% to 25%; however, after excluding tumors of low malignant potential or borderline tumors, and metastatic tumors to the ovary, largely from the GI tract, the true incidence of these tumors is much lower. Therefore, out of the 21,750 new cases of ovarian cancer estimated to be diagnosed in the United States, approximately 650 of these cases will be mucinous ovarian carcinomas. The rarity of these tumors is reflected in the low percentage of primary mucinous ovarian carcinomas represented in clinical trials. In fact, mucinous ovarian carcinomas represented 2–4% of enrolled patients in landmark GOG trials that helped establish current platinum-based treatment paradigms in epithelial ovarian cancer.
| Subtype | Frequency a |
|---|---|
| High-grade serous | 70% |
| Low-grade serous | 10% |
| Endometrioid | 10% |
| Clear cell | 5% |
| Mucinous | 3% |
| Other (Brenner, undifferentiated) | 2% |
a Frequencies for women in North American and Europe, frequencies differ in Asia.
Over 80% of mucinous ovarian carcinomas are diagnosed at an early stage, with disease confined to the ovary. The clinical presentation will typically be a young woman reporting pelvic pain or discomfort in the setting of a large (10–15 cm) unilateral ovarian mass. This contrasts with patients who are diagnosed with the more common type of epithelial ovarian cancer, high-grade serous ovarian carcinoma, which tends to present with advanced-stage disease due to the later onset of symptoms. Another feature distinguishing mucinous tumors from the high-grade serous subtype is that they are diagnosed more frequently in younger women, at a median age of 53 years, with 26% diagnosed in women younger than 44 years. In contrast, patients with high-grade serous ovarian carcinoma have a median age at diagnosis of 61 years, with less than 7% being younger than 44 years ( Table 5.2 ).
| Mucinous carcinoma | High-grade serous carcinoma | |
|---|---|---|
| Median age at diagnosis (years) | 53 | 61 |
| < 44 years at diagnosis (%) | 26 | 7 |
| Risk factors | Smoking | Germline BRCA mutations, nulliparity, early menarche, late menopause |
| Rate of bilaterality (%) | < 20 | > 50 |
| Stage I at diagnosis (%) | 65–83 | < 5 |
| Rate of lymph node metastasis for stage I disease (%) | 0 (Expansile) up to 30 (infiltrative) | 10 |
| Serum tumor markers | CEA or CA 19-9 | CA-125 |
| Rate of response to platinum-based chemotherapy (%) | 12.5–38.5 | 67.6–70 |
| Overall survival | ||
| Stage I, at 5 years (%) | > 90 | > 80 |
| Advanced stage (months) | 14.8 | 45.2 |
Women diagnosed with mucinous ovarian carcinoma have an overall better prognosis than those diagnosed with high-grade serous ovarian carcinoma. This is due to the fact that the majority of mucinous ovarian carcinomas are stage I at the time of diagnosis, in contrast to high-grade serous ovarian carcinomas, where stage I disease is rarely diagnosed (83% vs 4%). While early-stage mucinous ovarian carcinomas are associated with a 5-year overall survival exceeding 90%, patients with advanced-stage mucinous carcinomas tend to fare worse than their high-grade serous counterparts, with a median overall survival of less than 15 months. This is largely due to the relative resistance of mucinous ovarian carcinomas to platinum-based chemotherapy regimens ( Table 5.2 ).
Etiology/risk factors
It is now commonly believed that the majority of high-grade serous ovarian carcinomas or tubal carcinomas originate in the fimbriae of the tube and then exfoliate onto the surface of the adjacent ovary. In contrast, mucinous ovarian carcinoma appears to develop as a continuum, arising from benign epithelium, progressing to a mucinous borderline tumor before developing into an invasive malignancy. This is supported by the observation that invasive mucinous carcinomas often coexist with benign mucinous cystadenomas and/or borderline tumors within the same ovarian mass. Mucinous carcinomas can also arise from mature cystic teratomas or Brenner tumors. Histologically, the distinction between these different mucinous entities can be subtle, making an accurate pathologic diagnosis challenging.
Unlike high-grade serous ovarian carcinomas, for which nulliparity, early menarche, late menopause and germline BRCA1 or BRCA2 mutations are established risk factors, smoking is the only known risk factor associated with mucinous ovarian carcinomas ( Table 5.2 ). In a systematic review, current or prior history of smoking was associated with an increased risk for mucinous ovarian carcinoma (RR 2.1, 95% CI 1.7–2.1) but not the other epithelial subtypes of ovarian cancer. Current and former smokers also have worse cancer-specific survival compared with nonsmokers, with a hazard ratio of 1.9 (95% CI 1.01–3.65).
Genetic sequencing can help distinguish mucinous ovarian carcinomas from other epithelial subtypes and from extraovarian primaries. The vast majority of mucinous ovarian carcinomas harbor KRAS mutations (45%–71%), compared with only 10% of endometrioid ovarian carcinomas, 5% of high-grade serous ovarian carcinomas, and 0% of clear cell carcinomas. Notably, the same KRAS mutations have been found in the associated benign and borderline lesions, suggesting that this mutation is a founder event. While TP53 mutations are characteristic of high-grade serous ovarian carcinomas (> 96%), 56%–75% of mucinous ovarian carcinomas harbor these alterations. Additionally, amplification of ERBB2 (HER2) is also seen in 18%–35% of cases. Mutations in ERBB2 (HER2) and TP53 are only observed in the carcinoma component, suggesting that these alterations occur (develop) later in the process of malignant transformation.
Pathology
Gross
Majority of primary ovarian mucinous carcinomas are grossly large unilateral masses. This appearance is different from metastatic tumors involving the ovary, which are frequently bilateral; and the ovaries are small and multinodular with a bosselated surface. While the aforementioned appearance is typical, on occasion large unilateral tumors may represent metastasis from the gastrointestinal (GI) tract or pancreas rather than ovarian primaries; therefore, careful histologic evaluation is necessary.
Microscopic findings
Histologically primary ovarian mucinous carcinomas usually show a heterogeneous appearance with areas of cystadenoma, borderline tumor, and carcinoma. The majority of mucinous carcinomas are of intestinal type with the lining epithelium containing goblet cells. Mucinous ovarian carcinomas show two patterns of invasion, i.e., expansile type and infiltrative type. In the expansile pattern the glands are back to back with minimal or no intervening stroma, and this area should measure at least 5 mm in one dimension; importantly there is no stromal invasion or desmoplastic reaction around the glands ( Fig. 5.2 ). Infiltrative invasion is characterized by haphazard glands surrounded by desmoplastic stroma, measuring at least 5 mm in one linear dimension ( Fig. 5.3 ). If the infiltrative pattern of invasion is focal and measures < 5 mm, in a background of a mucinous borderline tumor, the term microinvasive carcinoma is used. The size criteria are based on the WHO Classification of Tumors of the Female Reproductive Organs. Of note, intestinal-type mucinous carcinomas may arise in a background of an ovarian teratoma and mimic metastasis due to extensive pseudomyxoma ovarii/peritonei.
A subset of mucinous carcinomas have mural nodules which may be composed of anaplastic carcinoma, or sarcoma. These two components have histologic overlap but the anaplastic carcinoma is diffusely positive for keratin markers, while negative in sarcoma. The presence of anaplastic carcinoma and sarcoma may be associated with an adverse outcome but some studies have not shown this to be the case in stage IA tumors; however, experience is limited. Sarcoma-like mural nodule, on the other hand, is a reactive proliferation composed of mitotically active spindle cells with numerous osteoclast-like giant cells, and as such is a benign proliferation that does not alter the outcome.
Ancillary testing
Primary ovarian mucinous tumors are usually positive for CK7 (diffuse) ( Fig. 5.4 ). CK20 ( Fig. 5.5 ), PAX-8 and ER show variable staining ranging from negative to focal or patchy staining but are not usually diffuse and strong. Markers of intestinal differentiation such as SATB2 are negative in the surface epithelial tumors, but can be positive in mucinous tumors arising in a background of teratoma. Distinction from metastasis can be challenging on histologic evaluation hence, doing a panel of immunostains is prudent.
Differential diagnosis
The differential diagnosis for ovarian mucinous carcinoma is quite diverse. Among other ovarian primaries, endometrioid carcinoma and clear cell carcinoma can show extensive mucinous differentiation and may be classified as mucinous carcinoma. Diffuse strong expression of ER and PR favors endometrioid carcinoma over mucinous carcinoma. Expression of HNF-1B and napsin-A would support clear cell carcinoma. Distinguishing metastatic mucinous ovarian carcinoma from a primary ovarian mucinous carcinoma can be particularly challenging. Histologic features such as dirty necrosis are typical of metastatic colonic adenocarcinoma, and the presence of pseudomyxoma ovarii/peritonei is almost always associated with an appendiceal primary. But in many cases there are overlapping histologic features and immunohistochemical (IHC) stains are necessary to make the correct diagnosis. The two most important markers are CK7 and CK20, the former being diffusely positive in ovarian mucinous carcinomas and the latter in tumors of the lower GI tract. However, right-sided colonic adenocarcinomas that are microsatellite unstable may be diffusely CK7 positive and negative for CK20, and most ovarian mucinous carcinomas can show focal CK20 staining; these factors must be considered when interpreting the stains. CDX2 is often positive in ovarian mucinous carcinomas which limits its utility in this differential. Other metastatic tumors that are morphologically virtually indistinguishable from ovarian mucinous carcinomas are tumors of the pancreaticobiliary tract and some cervical carcinomas. Pancreaticobiliary tract tumors show loss of DPC4 in about 50% of cases and this marker may facilitate the diagnosis as other markers are not specific in this differential. Unlike primary ovarian mucinous carcinomas, HPV-associated cervical carcinoma will be diffusely positive for p16 and show positivity for high risk HPV by in-situ hybridization. Despite the use of immunomarkers, in a subset of cases, distinguishing primary from metastatic tumors is not possible, and correlation with clinical and radiologic findings is necessary to determine the primary site of origin.
Molecular alterations
The most common molecular alterations in ovarian mucinous carcinoma, besides copy number loss, are mutations in CDKN2A , KRAS, and TP53 genes. HER2 gene amplification has been reported in approximately 25% of cases.
Diagnosis and workup
Differential diagnosis
It is challenging to accurately diagnose mucinous ovarian carcinomas. Ovarian metastases from extraovarian primary cancers account for the majority of mucinous carcinomas involving the ovary. In 2011, Zaino et al. published a landmark paper revealing that, in a prospective GOG trial, the vast majority of diagnosed primary mucinous ovarian carcinomas were, in fact, metastases from other sites. Less than a quarter (23%) of mucinous carcinomas involving the ovary will be primary ovarian cancers. The most common primary origin of these carcinomas will belong to the GI tract (45%), followed by pancreas (20%), cervix (13%), breast (8%), and uterus (3%). Given this, expert gynecologic pathology review is essential and additional clinical evaluation is needed to rule out metastases from nonovarian primaries. This includes upper endoscopy, colonoscopy, mammogram, computed tomography (CT) and/or positron emission tomography (PET) and serum tumors markers. This work-up should be completed in patients diagnosed with mucinous ovarian carcinomas, given the high frequency with which an extraovarian primary will be diagnosed.
A validated algorithm that uses size and laterality of the tumor can help confirm the diagnosis of primary versus metastatic mucinous tumors. Unilateral tumors greater than 10 cm in diameter have been shown to be an ovarian primary in over 80% of patients, while 88% of unilateral masses smaller than 10 cm and 94% of bilateral masses will be metastases ( Fig. 5.6 ). Primary mucinous ovarian carcinomas are greater in size, with a mean of 16–20 cm (range 5–48 cm) compared with metastatic mucinous carcinomas, which are 11–12 cm on average (range 2–24 cm). Given the broad differential diagnosis of these tumors, which includes GI malignancies, in addition to upper endoscopy and colonoscopy patients should undergo careful inspection of the peritoneal cavity, including the small and large bowel, by a gynecologic oncologist ( Fig. 5.7 ).
The origin of mucinous tumors, whether primary to the ovary or originating elsewhere, can be difficult to determine by histologic examination alone. Measuring the size and laterality of the tumor is helpful. Additional features that favor an ovarian primary include the following: the coexistence of borderline and/or benign mucinous component or teratoma, Brenner tumor, mural nodules or an expansile (confluent) pattern. Features that favor metastatic adenocarcinoma include the following: ovarian surface involvement, vascular invasion, hilar involvement, extensive infiltrative pattern, signet-ring cells, prominent desmoplastic response, small clusters of tumor cells within corpora lutea or albicantia, numerous pools of mucin dissecting the ovarian stroma, or a nodular pattern of invasion. ( Table 5.5 ).
| Primary mucinous ovarian cancer | Metastatic mucinous cancer to ovary |
|---|---|
| Unilateral | Bilateral |
| > 10 cm | < 10 cm |
| Intact ovarian capsule | Ovarian surface involvement |
| Absence of vascular invasion, hilar involvement and extensive infiltration | Presence of vascular invasion, hilar involvement and extensive infiltration |
| IHC Markers: +CK7, + CK20 | IHC Markers: Colorectal primary: -CK7, + CK20 Breast primary: + CK7, − CK20 Cervical primary: + p16 Endometrial primary: + ER, + PR |
IHC markers can help narrow the diagnosis. While immunostaining with PAX8 is typical for tumors of gynecologic origin, PAX8 staining is usually focal and weak in primary mucinous ovarian carcinomas. Meanwhile, colorectal tumors tend to express SATB2, although mucinous tumors arising from mature teratomas can also stain diffusely positive for SATB2. Additionally, primary mucinous ovarian carcinomas stain positive for CK7 and CK20, while colorectal primaries tend only to express CK20, and breast primaries are usually CK7 positive and CK20 negative. Unlike endometrioid tumors of gynecologic origin and primary breast cancers, mucinous ovarian carcinomas do not express estrogen or progesterone receptors (ERs or PRs). HPV in situ hybridization and p16 immunostaining can be helpful in diagnosing a cervical primary, as p16 staining is usually negative in primary mucinous ovarian carcinomas. While the application of massively parallel sequencing is not yet widely available in the clinical setting, some have advocated a targeted gene panel with select IHC to narrow the differential diagnosis.
Signs and symptoms
In the case of primary mucinous ovarian carcinomas, the tumors tend to be large and unilateral. Most patients present with pelvic pain or discomfort, prompting medical evaluation. Patients with advanced-stage disease often have omental caking and peritoneal spread of disease in the pelvis and upper abdomen, and report increasing abdominal girth, bloating, early satiety and weight loss symptoms, similar to those experienced by patients with high-grade serous ovarian carcinomas.
Physical exam findings
In the majority of patients with primary mucinous ovarian carcinomas, abdominal and pelvic examination will reveal a large abdominopelvic mass. Laterality cannot always be determined based on examination alone, given the large size of these tumors. A thorough physical exam, including a pelvic exam to rule out a cervical or uterine mass, and a rectovaginal exam, should be performed to help determine extent of disease and to assess for a rectal primary. A breast examination should be considered if clinically indicated, as metastatic breast cancer can present with ovarian metastases ( Table 5.3 ).
| Physical exam | Complete physical exam with pelvic and rectovaginal examination Consider breast examination if indicated |
| Imaging | CT chest/abdomen/pelvis with contrast MRI pelvis if indicated PET CT (optional) Obtain endoscopy and colonoscopy pre or post operatively If breast etiology suspected, obtain mammogram as clinically indicated |
| Tumor markers | CEA, CA 19-9, CA-125 |
Tumor markers
Approximately a third of all ovarian cancers will have elevated carcinoembryonic antigen (CEA) levels; however, CEA is more likely to be elevated in mucinous ovarian carcinomas specifically. CEA is an oncofetal antigen that is found in embryonic and fetal tissue and disappears almost completely after birth; however, small amounts may persist in the adult colon. CEA may also be elevated in colon and pancreatic cancer and benign diseases involving the liver, lung and GI tract. CA-125, while elevated in 85% of high-grade serous carcinomas, is only elevated in 12% of mucinous ovarian carcinomas, with a mean CA-125 of 44.2 U/mL compared with 235 U/mL for high-grade serous carcinomas. Given the specificity for mucinous carcinomas, CEA is a more useful marker preoperatively and throughout the patient’s postoperative or treatment course. Carbohydrate antigen 19-9 (CA 19-9) is also more likely to be elevated in mucinous ovarian carcinomas and may help distinguish benign mucinous neoplasms from invasive carcinoma.
Imaging studies
CT scan of the chest, abdomen and pelvis with contrast is helpful in the primary work-up of a suspicious adnexal mass, to evaluate the lesion and radiologic extent of disease. While transvaginal ultrasound may be obtained as part of the initial evaluation of pelvic pain or when an adnexal mass is palpated on pelvic examination, ultrasound is insufficient to capture the entire size of these lesions and will not rule out the possibility of metastatic disease. MRI may be used to further characterize the tumor and its relationship with key structures of the pelvis ( Fig. 5.1 ); however, CT remains the preferred imaging modality. PET scans can be considered if nonovarian metastatic mucinous carcinoma is suspected; however, we do not routinely obtain PET scans in our practice.
Diagnostic testing
In addition to obtaining a detailed clinical history, a complete physical examination including a pelvic exam and rectovaginal exam should be performed. Imaging for pelvic masses typically begins with transvaginal ultrasound, given the wide availability of this modality. However, if there is suspicion for malignancy, a CT scan, preferably of the chest, abdomen and pelvis, should be obtained to assess for metastatic disease and resectability. Relevant tumor markers, including CA-125, CEA and CA 19-9, should be obtained ( Table 5.3 ). While a primary mucinous ovarian carcinoma may be suspected based on clinical presentation and imaging findings, the diagnosis is made on pathological evaluation following surgical resection. Given the large size of these tumors, it may be difficult to provide a diagnosis of invasive carcinoma on frozen section, making it a challenge to counsel a patient on management until the final pathology is clarified. Furthermore, distinguishing a primary mucinous ovarian carcinoma from metastatic mucinous ovarian carcinoma adds another layer of complexity, given the morphologic similarity of these tumors. In addition to endoscopic studies to rule out a GI primary, including upper endoscopy and colonoscopy, IHC studies can help narrow the diagnosis (see Sections “ Pathology ” and “ Differential diagnosis ”).
Staging system
Mucinous ovarian carcinomas are staged according to the International Federation of Gynecology and Obstetrics (FIGO) ovarian cancer staging system ( Table 5.4 ). Surgical staging of mucinous ovarian carcinomas follows epithelial ovarian cancer guidelines and includes removal of the uterus, cervix, both tubes and ovaries, and obtaining peritoneal cytology, multiple peritoneal biopsies, and an omental biopsy or omentectomy. For patients desiring fertility preservation, the unaffected contralateral ovary and uterus may be retained if disease is confined to a single ovary; however, staging with or without lymphadenectomy should still be performed. The current staging guidelines include comprehensive lymphadenectomy; however, given the rarity of occult nodal involvement (< 2%) in mucinous ovarian carcinomas, lymph node sampling can be omitted. While many institutions grade mucinous ovarian carcinomas according to the FIGO system, it is important to note that there is no standardized grading system for primary mucinous ovarian carcinomas.
