Neuroendocrine tumors
Epidemiology
Incidence and mortality
Neuroendocrine carcinomas (NEC) of the endometrium, which includes both small-cell and large-cell variants, are rare, highly aggressive tumors that account for approximately 0.8% of endometrial carcinomas. They generally present with advanced stage disease. The literature consists of only case reports and small case series. In one case series, 32% of patients presented with early (stages I–II) disease, whereas 68% presented with advanced stage disease, with a predominance of incidence in the sixth and seventh decades of life. Another case series reports 26%, 5%, 36%, and 33% of patients present with stages I, II, III, and IV, respectively.
Median survival has been reported as 17 months with a 5-year survival of 38.3%. Stage for stage, 5-year survival for uterine neuroendocrine tumors (NETs) is significantly worse than survival for women with endometrioid adenocarcinoma of the uterus ( Table 13.1 ). Mean survival ranges from 12 months for advanced stage disease to 22 months for early stage disease. In a cohort of 25 patients, seven attained 5-year survival, of which three patients had stage I disease and four had stage III disease.
| Stage | Neuroendocrine carcinoma | Endometrioid carcinoma |
|---|---|---|
| IA | 76.4% | 85.4% |
| IB | 65.0% | 72.7% |
| II | 37.5% | 67.1% |
| IIIA | 16.1% | 58.2% |
| IIIB | 45.0% | 46.4% |
| IIIC | 34.2% | 53.1% |
| IVA | 0.0% | 28.8% |
| IVB | 15.4% | 27.7% |
Etiology and risk factors
NEC of the endometrium are a histologically distinct subtype of endometrial carcinoma with a propensity of metastatic infiltration and an unfavorable prognosis. They tend to form bulky, intraluminal masses with deep myometrial involvement. They are subdivided into small- and large-cell tumors according to nuclear size. Endometrial NEC are often combined with other epithelial neoplasms, with 50%–80% of cases admixed with FIGO grade 1 or 2 endometrioid carcinoma.
Based on the largest population-based study ( N = 364) from the National Cancer Database (NCDB), compared to women with poorly differentiated endometrial cancer, women with NEC were more often black (13.7% neuroendocrine vs. 10.7% poorly differentiated endometrial cancer) and Hispanic (8.0% neuroendocrine vs. 5.3% poorly differentiated endometrial cancer). Additionally, as with other endometrial carcinomas, postmenopausal status is a known risk factor.
Pathology
Gross description
Grossly the tumors are large, endometrial based and typically with deep myometrial invasion. Tumor size ranges from 0.8 to 12 cm with a median tumor size of 6.0 cm. No gross pathognomonic features have been described.
Microscopic description
Small-cell neuroendocrine carcinoma (SCNECa) resembles that of pulmonary small-cell carcinoma, with “salt and pepper chromatin” and scant cytoplasm with nuclear molding. Large-cell neuroendocrine carcinoma (LCNECa) typically have neuroendocrine growth pattern, such as organoid/nesting ( Fig. 13.1A ), trabecular or corded. LCNECa characteristically have polygonal cells with abundant cytoplasm and prominent nucleoli ( Fig. 13.1B ). The tumors frequently have central geographic necrosis, numerous mitotic figures and apoptotic bodies. Tumors may show pure SCNECa or LCNECa or may show mixed small and large-cell morphology. NEC can also be mixed with other histotypes, including endometrioid adenocarcinomas, serous carcinomas and carcinosarcomas.
Immunohistochemistry
Epithelial markers such as a pankeratin cocktail should be positive, while rare tumors that are negative for pancytokeratin cocktail may have expression of cytokeratin 18. In some cases, several markers may be necessary to establish epithelial origin. NEC are positive for at least one neuroendocrine marker in at least 10% of the tumor cells, including synaptophysin, chromogranin, and CD56. It should be noted though that non-NECs, such as an endometrioid FIGO grade 3 or undifferentiated carcinoma may show staining for any one of the neuroendocrine markers, and that interpretation of IHC should be in the context of the overall morphology. In some cases where the morphology and or IHC are not characteristic, a diagnosis of neuroendocrine differentiation or features may be used by the pathologist. p16 can be positive in a patchy or diffuse pattern, with no evidence of infection with high-risk HPV.
Differential diagnosis
The differential diagnosis includes NEC from other primary sites, such as extension from the cervix or rarely metastases from other sites. In large tumors that involve the cervix and endometrium, establishing the primary site requires immunohistochemical staining, as the tumors are histologically similar. p16, which is usually a surrogate marker of HPV infection, can be diffusely expressed by non-HPV associated pathways and is not helpful in this distinction. Therefore, testing for HR-HPV is usually required and a negative result would favor endometrial primary in the correct clinical setting. High-grade carcinomas with a solid or nested growth pattern, such as FIGO grade 3 endometrioid carcinoma or serous carcinoma, may warrant consideration of a NEC. While neuroendocrine markers may be exhibited in these tumors, IHC is focal and less than 10% of tumor cells. Undifferentiated/de-differentiated carcinoma is also a consideration, usually lacking a nested architecture and with a diffuse, sheet-like pattern of growth. Like other NEC, focal staining for neuroendocrine markers may be seen but should be less than 10% of tumor cells.
Molecular findings
Approximately 40% of a series of endometrial NEC showed loss of expression of mismatch repair proteins, with the most common pattern being loss of MLH1/PMS2. Molecular analysis on a separate series shows that endometrial NEC also classify into The Cancer Genome Atlas (TCGA) groups, with POLE- mutated/ultramutated (7%), MSI/hypermuted (43%) and TP53 mutated/copy number high (14%). Some SCNECa may show similar molecular profiles to those expected in typical pulmonary SCNECa ( RB1 deletion and TP53 mutations).
Diagnosis and workup
The differential diagnosis for NEC is broad and includes undifferentiated/dedifferentiated endometrial carcinoma, carcinosarcoma, poorly differentiated endometrioid carcinoma, serous carcinoma with a solid growth pattern and massive necrosis, primitive neuroectodermal tumors (PNETs), and carcinoid tumor. Additionally, NEC is often associated with tumors of other histologic types, including complex atypical hyperplasia, endometrioid carcinoma, carcinosarcoma, or serous carcinoma.
In a 25 patient series, the most common presenting symptom was vaginal bleeding (15 of 25 patients) with other less common presenting symptoms including abdominal pain, symptoms related to metastasis, and abnormal Pap test. Additional common presenting symptoms include postmenopausal bleeding and pelvic pain. Of note, in very rare cases, NEC can also present with signs of paraneoplastic disease and/or Cushing’s syndrome.
Workup begins with a complete physical exam including pelvic and rectovaginal examinations ( Table 13.2 ). The most common physical exam findings include abnormal vaginal bleeding, abdominal pain, and an enlarged uterus and/or a palpable pelvic mass. There is not a typical tumor marker that is elevated in NEC. CA-125 may be elevated in the setting of a mixed tumor, but this marker is not specific to NETs.
| Component of workup | Specific tests/considerations |
|---|---|
| Physical exam | Pelvic and abdominal exams |
| Tumor markers | Consider CA-125 |
| Imaging |
|
| Diagnostic tests | Endometrial biopsy or dilation and curettage |
Data on the optimal pretreatment imaging are limited. Consideration can be given to obtaining a transvaginal pelvic ultrasound and a chest radiograph. If there is an abnormality on chest radiograph, a chest computed tomography (CT) can be obtained to further evaluate. If there is concern for high-grade carcinoma (e.g., if already diagnosed as high-grade tumor by an endometrial biopsy (EMB)), it is reasonable to obtain a CT of the chest, abdomen, and pelvis to evaluate for metastatic disease. Other initial imaging may be obtained based on symptoms and/or clinical concern for specific metastatic sites.
In some cases, an MRI abdomen and pelvis may be obtained. MRI findings are similar to other poorly differentiated carcinomas, with an ill-defined endometrial-myometrial border and heterogeneity of the mass indicating necrosis and hemorrhage. Consideration can also be given to positron emission tomography (PET).
Similar to other endometrial cancers, the gold standard for initial diagnosis is by pathologic diagnosis obtained from either an EMB or endometrial curettings obtained from a dilation and curettage (D&C).
Staging system
Uterine NEC is surgically staged utilizing the 2017 International Federation of Gynecology and Obstetrics (FIGO)/ Tumor, Node, Metastasis (TNM) system used for endometrial carcinoma ( Table 13.3 ).
| Stage | Description |
|---|---|
| I | Tumor confined to uterus IA: < 50% myometrial invasion IB: > 50% myometrial invasion |
| II | Involvement of cervical stroma |
| III | Involvement of serosa, adnexa, vagina, parametrium, or locoregional lymph nodes IIIA: serosa and/or adnexa IIIB: vagina or parametrium IIIC: locoregional lymph nodes IIIC1: pelvic lymph nodes IIIC2: para-aortic lymph nodes |
| IV | Involvement of bladder/bowel mucosa, or distant metastases IVA: bladder/bowel mucosa IVB: distant metastases |
Prognostic factors
Factors associated with death in a NCDB series include: diagnosis at 80 years of age or older, stage (diagnosis at stage II or higher compared to stage I), and not receiving adjuvant chemotherapy. Women over the age of 80 have a hazard ratio for death of 2.4 compared to women less than 50 years old and patients who receive chemotherapy have a hazard ratio for death of 0.36 compared to those who do not get chemotherapy. In a smaller study of 42 patients, performance status, FIGO stage, surgery, and histologic subtype (pure type with worse prognosis than mixed type) were significant prognostic factors; in a multivariate analysis, only surgery and histologic subtypes remained significant. There was also a better prognosis in complete surgery cases (versus no surgery or incomplete surgery).
Prognostic factors for small-cell NETs in general (not specific to endometrial origin) include early detection and prompt surgical and adjuvant therapy.
Treatment of primary disease
Standard treatment guidelines do not exist given the rarity of NEC and lack of prospective data to guide therapy. Treatment regimens are largely based on traditional treatments of endometrial cancer and small-cell lung cancer. The majority of reported treatment regimens in the literature include surgical resection and adjuvant therapy with platinum-based chemotherapy with or without radiation therapy. Treatment usually begins with surgery for women without metastatic disease ( Fig. 13.2 ). Surgery generally includes hysterectomy, bilateral salpingo-oophorectomy (BSO), and lymph node assessment. For younger women, fertility preserving surgery can be considered on an individual basis but most patients should undergo BSO.
Adjuvant treatment is reasonable to consider given aggressive nature of this cancer type with high risk of recurrence and metastasis. In the largest reported series of cases, all patients underwent surgery, 60% received chemotherapy, and 28% received radiation. Compared to women with poorly differentiated endometrial carcinoma, those with NEC were more likely to receive adjuvant chemotherapy, likely owing to the tendency to present with later stage disease. In another series, all patients underwent surgery and 80% received adjuvant therapy. All patients undergoing surgery with the majority also receiving adjuvant chemotherapy is consistent with smaller series and case studies.
The most common reported chemotherapy regimens include cisplatin and etoposide. The following dosage regimen has been reported in the literature and is reasonable to consider for adjuvant treatment: cisplatin (80 mg/m 2 on day 1) and etoposide (80–120 mg/m 2 on days 1–3) over a 3-week interval. Of note, adjuvant treatment regimens used for NEC are largely based on data from small-cell lung cancer and neuroendocrine cervical cancer. A case series of SCNECa of the cervix demonstrated 83% recurrence-free survival (RFS) at 3 years with use of adjuvant etoposide and cisplatin. Indications for neoadjuvant chemotherapy are similar to those in other endometrial cancer types, and may be considered in the setting of advanced stage disease, though its use is rare in the reported literature.
In conclusion, there are no evidence-based therapeutic regimens for NEC due to its rarity, but surgery followed by adjuvant chemotherapy (most commonly cisplatin and etoposide) has been used with several patients obtaining long-term survival in case reports.
Surveillance for recurrence
Based on recommendations for surveillance of endometrial carcinoma, it is reasonable to consider physical exam every 3–6 months for 2–3 years and then every 6 months or annually with thorough review of systems and tumor markers (if initially elevated) at each visit, as well as imaging if clinically indicated. Evidence-based recommendations for surveillance are lacking.
Survival and patterns of failure
A large case series reports 5-year progression-free survival (PFS) and overall survival (OS). Five-year PFS is 81.8%, 50%, 32.5%, and 14.3% for FIGO stages I, II, III, and IV, respectively. OS is 88.9%, 100%, 46.7%, and 21.4% for FIGO stages I, II, III, and IV, respectively. Recurrence occurred in 50% of cases with a higher rate of hematogenous recurrence compared to lymphogenous or local recurrence.
Treatment of recurrent disease
There are no standard regimens due to rarity of the disease ( Table 13.4 ). One report in the literature describes the use of salvage chemotherapy with cisplatin (75 mg/m 2 ) and ifosfamide (5 g/m 2 ). Second-line therapies can be considered based on cervical and lung cancer data: cyclophosphamide/doxorubicin/vincristine (CAV), irinotecan/platinum (IP), or topotecan.
| Treatment options for recurrent disease | Specific regimens/agents |
|---|---|
| Salvage chemotherapy | Cisplatin and ifosfamide |
| Second-line therapies (used in cervical and lung cancer) | Cyclophosphamide/doxorubicin/vincristine (CAV), irinotecan/platinum (IP), or topotecan |
| Targeted therapies | PI3K and MEK inhibitors |
| Immunotherapy | Nivolumab (PD-1 receptor inhibitor) |
Based on preliminary data from NEC of the cervix, PI3K and MEK inhibitors are targeted therapies that may have therapeutic potential for treatment of endometrial NEC. One study looked at a newly established neuroendocrine cervical carcinoma cell line and found that when etoposide and cisplatin were combined with a PI3K inhibitor (which functions in the mTOR pathway), the growth of tumor cells was significantly reduced. Another study examined a MEK inhibitor, trametinib, which functions in the RAS pathway and is FDA-approved for treatment of unresectable or metastatic melanoma in patients with BRAF mutations, in the treatment of a patient with recurrent cervical neuroendocrine carcinoma with a KRAS mutation; she has been on trametinib for 8 cycles and is without evidence of disease. Based on data from pulmonary small-cell tumors, a number of other targeted therapies have been studied including angiogenesis inhibitors, mTOR inhibitors, and PARP inhibitors.
There is a case report evaluating the use of a programmed cell death 1 (PD-1) receptor inhibitor, nivolumab, in the treatment of a patient with recurrent, metastatic SCNECa of the cervix. The patient experienced a complete response and remained free of disease 4 months after stopping treatment. Further research is needed to see if immunotherapy is a reasonable approach to recurrent NEC of the cervix and/or endometrium.
Yolk sac tumors
Epidemiology
Incidence and mortality
Yolk sac tumors (YSTs), also known as endodermal sinus tumors, are malignant germ cell tumors that secrete alpha-fetoprotein (AFP). The most common origin is gonadal, but 20% of cases arise from various extra-gonadal sites (mediastinum, retroperitoneum, sacrococcygeal region, pelvis, liver, stomach, vulva, vagina, and cervix). The most common extra-gonadal site in the female genital tract is the vagina.
Primary YSTs of the endometrium are very rare with fewer than 20 case reports in the literature. The median survival for endometrial YSTs is 28 months (range, 8–72 months). In review of the existing case reports, only 3 (out of 19) reported mortality ; however, periods of follow up were variable. Five-year survival rates for ovarian YSTs, which are much more common than those of endometrial origin, for stages I, II, III, and IV are 95%, 75%, 30%, and 25%, respectively.
Etiology and risk factors
Four possible mechanisms for germ cell tumors in the endometrium have been proposed: aberrant lateral migration of primordial germ cells during embryogenesis, metastases from an occult focus in the ovary, tumor origination from residual fetal tissue after incomplete abortion, or aberrant differentiation of somatic cells. Given the rarity of this tumor, specific risk factors have not been determined.
Pathology
Extragonadal YSTs are exceedingly rare. YST, in comparison to other germ cell tumors, displays diverse histologic patterns. Due to this variability, YSTs can commonly mimic somatic tumors, posing a significant diagnostic challenge, even more so in extragonadal sites.
Gross description
Due to the rarity of these tumors, there is a paucity of information regarding the gross appearance of uterine YSTs. YSTs have been described as tan-yellow, soft, friable and hemorrhagic. Most patients present with abnormal vaginal bleeding and some with a uterine mass. Approximately 70% of uterine YSTs present with FIGO stage II or higher.
Microscopic description
YSTs have various histologic patterns, including microcystic/reticular, glandular, solid, papillary and hepatoid ( Fig. 13.3 ). Approximately one-third of cases have only a single identifiable histologic pattern, while the rest show an admixture of patterns with a predominant pattern that is typically microcystic/reticular, glandular or solid. Schiller–Duval bodies, while useful as a histologic clue for YSTs, are typically rare, both in the percentage of cases they are present in, as well as within the tumors themselves. Approximately 70% of uterine YSTs present with an associated somatic component that may represent < 10% to 90% of the total sampled tumor, including the following histotypes: complex atypical hyperplasia, serous carcinoma, carcinosarcoma, and mixed high-grade carcinomas.
Immunohistochemistry
YSTs are typically positive for SALL4, AFP, and Glypican-3, while they show focal or no staining for cytokeratin 7 and PAX-8 ( Fig. 13.4 ). YSTs, especially those with glandular pattern, may express markers typically associated with a somatic tumor, including villin, CDX2 and cytokeratin 20, and pose a diagnostic pitfall.
Molecular findings
There are very few studies evaluating the molecular characteristics of uterine YSTs. A case report of a pure uterine YST showed allelic copy gains at STR loci on chromosomes 2, 3, 4, 5, 8, 13, and 15, with copy loss at chromosome 1, with no point mutations or indels identified in a 155 gene panel. Overall, these findings are suggestive of a germ cell origin for this one case, rather than a somatic stem cell. On the other hand, a more recent study that performed next generation sequencing in a case of endometrial mixed clear cell carcinoma and YSTs showed ARID1A and TP53 mutations in both components, supporting a somatic derivation for the YST component. It is likely that some uterine YSTs are of germ cell origin and others are somatically derived.
Differential diagnosis
The differential diagnosis includes somatic carcinomas in the uterus that may show glandular or clear cell changes, including endometrioid adenocarcinoma, clear cell carcinoma, serous carcinoma or carcinosarcomas. Careful attention to the histology and awareness of the varied histology of YSTs can at the least trigger an IHC panel to rule in or rule out the presence of a YST component. Another important pitfall includes colorectal adenocarcinoma as YSTs can express CK20, CDX2, villin, which are markers of intestinal differentiation.
Diagnosis and workup
Histologically, YSTs are often confused with endometrial clear cell carcinoma. However, patients with clear cell carcinomas are typically older. Other conditions that can cause an elevated AFP include hepatitis, extensive hepatic metastasis from an alternative primary tumor, and various malignancies of the gastrointestinal (GI) tract.
The most common symptom associated with endometrial YSTs is abnormal uterine bleeding or postmenopausal bleeding. Additional presenting symptoms that have been reported in the literature include abdominal and pelvic pain and watery vaginal discharge.
Workup begins with a complete physical exam including pelvic and rectovaginal examinations ( Table 13.5 ). Physical exam findings include bulky/enlarged uterus, prolapsing uterine tumor, and pelvic/abdominal fullness. A minority of cases in the literature also report unremarkable pelvic exams.
| Component of workup | Specific tests/considerations |
|---|---|
| Physical exam | Pelvic and abdominal exams |
| Tumor markers | Alpha-fetoprotein (AFP); consider β-hCG, CA-125, CEA |
| Imaging |
|
| Diagnostic tests | Endometrial biopsy or dilation and curettage |
YSTs commonly present with elevated AFP. In case reports in the literature, AFP at time of diagnosis ranges from 291.9 to 38,366 ng/mL (normal reference range 1.01–7.10 ng/mL). There are two case reports with normal AFP. If there is a mixed germ cell tumor, other tumor markers including beta-human chorionic gonadotropin (hCG), CA-125, and CEA may be elevated.
The most commonly used imaging modalities include pelvic ultrasound followed by CT and magnetic resonance imaging (MRI). A minority of studies assessed for metastatic disease with chest imaging in the absence of pulmonary symptoms. Pelvic imaging commonly shows a thickened endometrial stripe and/or an intrauterine mass.
Similar to other endometrial cancers, the gold standard for initial diagnosis is by pathologic diagnosis obtained from either an EMB or endometrial curettings obtained from a D&C.
Staging system
Similar to uterine NEC described above, uterine YSTs are surgically staged utilizing the 2017 FIGO/TNM system used for endometrial carcinoma ( Table 13.3 ).
Prognostic factors
There are too few case reports of endometrial YSTs to determine prognostic factors with any degree of certainty. However, one can extrapolate from ovarian YSTs for prognostic factors, with the knowledge that the tumors may behave in slightly different ways.
A series of ovarian YSTs found that patients with early stage disease, ascites volume less than 100 cc, and residual tumor less than 1 cm after surgery had a more favorable prognosis. Another retrospective review found the following factors to be associated with a more favorable prognosis: early stage, size of residual tumor less than 2 cm, absence of ascites or less than 100 cc in volume, and use of chemotherapy with cisplatin (versus without). This study did not find pre-operative AFP level or lymph node status to be prognostic indicators. However, both studies are limited by small sample size.
Treatment of primary disease
Treatment has not been standardized due to the rarity of YSTs of the endometrium. The general consensus regarding the treatment approach for YSTs of the endometrium is a combination of surgery and chemotherapy ; this has largely been extrapolated from germ cell tumors in other locations ( Fig. 13.5 ).
