Historical perspective

Sarcomas (including mesenchymal and mixed epithelial–mesenchymal malignancies) of the vulva, vagina, cervix, uterus, and ovaries account for less than 1.5% of the cancers of these organs. Classification of these cancers was a taxonomic dilemma until Ober, in 1959,¹ proposed a classification which Kempson and Bari² revised in 1970 and the World Health Organization³ and the College of American Pathologist⁴ reclassified in 2003 (Table 1). Uterine carcinosarcomas are not classified as a uterine sarcoma any longer and should be classified as a metaplastic carcinoma of the uterus⁵ but is discussed in this chapter.

Low-grade endometrial sarcoma is distinguished from benign endometrial stromal nodule by infiltration into the surrounding myometrium and/or lymphovascular invasion. Minor marginal irregularity in the form of tongues <3 mm (up to 3) is allowable for an endometrial stromal nodule. This protocol does not apply to endometrial stromal nodule.

Data from Tavassoli³ and Otis.⁴

Incidence and epidemiology

The most common site for sarcoma in the female pelvis is the uterus comprising only 4–9% of uterine cancers, with an annual incidence rate of less than 20 per million females.^{6, 7} Overall incidence for Blacks is twice that of Whites, but there were no differences in survival for women receiving similar therapy.^{6, 7} Risk of carcinosarcoma increases sharply with age. Incidence rate per million women per year is 8.2 for carcinosarcomas, 6.4 for leiomyosarcomas (LMSs), 1.8 for endometrial stromal sarcomas (ESSs), and 0.7 for unclassified sarcomas.^{6, 7} Authors have reported that carcinosarcoma and LMS each account for 35–40% of uterine sarcomas, ESS accounting for 10–15%, and other sarcomas comprising 5%.⁷

Risk factors

Epidemiologic risk factors for uterine sarcoma are undefined except for radiation exposure^{8, 9} and previous tamoxifen use.^{10, 11}

Pathology

The uterus is the most common site of gynecologic sarcomas arising from the endometrium only (ESS), the myometrium only (LMS), or contributions from both (carcinosarcomas).^{1, 12} The homologous tumors include carcinoma plus a sarcoma indigenous to the uterus, while the heterologous tumors includes a sarcoma resembling tissue from some extrauterine source (bone, cartilage, striated muscle). Müllerian adenosarcomas are mixed müllerian tumors composed of malignant stroma and benign epithelium.^{3, 13, 14}

Endometrial stromal lesions

Endometrial stromal nodules (ESN) are rare. They are characterized by a well-defined noninfiltrating border without evidence of myometrial or vascular invasion. Two-thirds are found as isolated lesions within the myometrium with no apparent connection to the endometrium.³

ESS is low grade with metastatic potential. Like ESN, they are composed of uniform cells that mimic proliferative endometrium. However, they exhibit myometrial and/or vascular invasion.³ Histologically, they are characterized by densely uniform stromal cells with minimal cellular pleomorphism, mild nuclear atypia, and rare mitotic figures. Of note, an isolated finding of increased mitotic figures does not confer an adverse prognosis in an otherwise typical low-grade ESS.¹⁵

The diagnosis of ESS may be complicated by variant morphologic features (Table 1).³ In tumors with focal smooth muscle differentiation, the tumor is categorized as ESS if the smooth muscle component involves <30% of the total tumor volume. Tumors composed of a larger smooth muscle component are designated as mixed endometrial stromal and smooth muscle tumors.^3,16

Undifferentiated endometrial sarcoma (UES—previously referred to as high-grade ESSs) is characterized by marked cytologic atypia, nuclear pleomorphism, high mitotic activity, and extensive invasion. In addition, UES usually show destructive myometrial invasion.³

Molecular and genetic alterations

The following molecular features are characteristic of ESS and are also found in ESN: The majority are immunoreactive for the estrogen receptor (ER) and progesterone receptor (PR). They are typically immunohistochemically positive for CD10 and negative for desmin and h-caldesmon and loss of heterozygosity of PTEN, and deregulation of the Wnt signaling pathway.^{5, 17, 18}

UES shows increased staining for proliferation markers (Ki67, p16, and p53) and does not generally exhibit immunoreactivity against ER, PR, desmin, or smooth muscle antigen (SMA).⁵ UES also expresses the receptor tyrosine kinase CD117 (c-KIT),¹⁹ cyclin D1,²⁰ and human epidermal growth factor receptor 2 (HER2 or ERBB2), which are not typically found in ESS.⁵

Gene rearrangements have been described and validated in patients with ESSs with at least 75% having a gene rearrangement.^{21, 22} The t(7;17) translocation resulting in the JAZF1-SUZ12 gene fusion is the most common translocation found in approximately 35–50% of ESSs.²² Other noted gene fusions are JAZF1-PHF1, EPC1-PHF1, JAZF1 only, and PHF1 only.^21–25

UESs lack these JAFZ1-based rearrangements but instead appear to frequently harbor the YWHAE-FAM22A/B genetic fusion,^26–29 which may be specific to these tumors as this rearrangement was not detected in 827 other cases representing 55 tumor types.²⁷

Patterns of spread

Uterine sarcomas are spread by lymphatic, hematogenous, local extension, and peritoneal dissemination.^30–38

Rose et al.³⁰ studied the autopsy findings of 73 patients with uterine sarcoma, including 43 patients with carcinosarcoma, 19 with LMS, 9 with ESS, and 2 with endolymphatic stromal myosis. The peritoneal cavity and omentum were the most frequently involved sites (59%), followed by lung (52%), pelvic (41%) and paraaortic (38%) lymph nodes, and liver parenchyma (34%). Of note, the presence of lung metastasis was often a sole metastatic site.

Lymph node metastasis in adult soft-tissue sarcomas is <3%, with some variation among histologic subtypes.³⁹ The risk of lymph metastasis in LMS overall was approximately 6.4% in a series of 357 patients.⁴⁰ However, the rate of occult lymph node metastasis in clinically normal nodes and disease clinically confined to the uterus was only 3.5% among 57 surgically staged patients with LMS.³⁵ Carcinosarcomas have a higher rate of both overall (25%)⁴¹ and occult (18%)³⁵ lymph node metastasis. The rate of overall and occult lymph node metastasis in ESS is 16% and 6%, respectively.⁴² Deep myometrial invasion and extensive lymph–vascular space invasion (LVSI) further increase the risk of occult metastasis.^{32, 43} The rate of lymph node metastasis in adenosarcomas is approximately 3%.⁴⁴

Clinical profile

Tumor biomarkers

Goto et al. reported that serum lactate dehydrogenase (LDH) was elevated in a small series of LMS and was useful in combination with dynamic MRI. The positive predictive value was 91% using the combined assessment compared to 39% for LDH alone and 71% for MRI alone.⁵¹

Imaging studies

ESS has a nonspecific appearance on ultrasound.⁵² The characteristic pattern of ESS consists of worm-like tumor projections along the vessels or ligaments, which are best visualized on MRI with diffuse-weighted imaging.⁵³ There are few studies describing the characteristic appearance of UES.⁵⁴

Kurjak et al.⁵⁵ evaluated the role of transvaginal color Doppler ultrasonography in differentiating uterine sarcomas from leiomyomas. Computed tomography will identify extrauterine spread, and magnetic resonance imaging can assess the depth of myometrial invasion. Imaging studies are performed preoperatively to characterize the uterine mass and evaluate for lymph node involvement and other metastases but cannot reliably differentiate between a uterine sarcoma and other uterine findings. There are few data suggesting the best choice of imaging.^{54, 56} The FDG-PET showed a better detection rate than the abdominal CT scan for extrapelvic metastatic lesions.^57–59

Diagnosis

Of patients with uterine sarcomas, 75–95% present with abnormal vaginal bleeding.^{60, 61} Pelvic pain, discharge, and aborting tissue occur frequently. Endometrial biopsy may confirm carcinosarcoma in the majority of cases⁶²; however, LMS and ESS are missed in at least 40% and 20% of cases, respectively.^{34, 63, 64}

TNM and FIGO staging classification

Uterine sarcomas require surgical staging. The International Federation of Gynecology and Obstetrics (FIGO) did not have a sarcoma-specific system until recently and the Endometrial Staging criteria were applied until 2009. FIGO devised a uterine sarcoma-specific staging system in 2009 for LMS, EES, and UES and another specific for adenosarcoma (Tables 2 and 3).

Data from Otis⁴ and D’Angelo and Prat.⁵

Modified from Otis⁴ and D’Angelo and Prat.⁵

Carcinosarcomas are to be staged using the 2009 revised FIGO Staging for endometrial carcinomas (see Chapter 103).

Prognostic factors and prognosis

Prognostic factors differ for the three major types of uterine sarcomas. Major and colleagues reported the GOG clinicopathologic study of clinical stage I and II uterine sarcoma, which included 59 patients with LMS and 301 patients with carcinosarcoma.³⁵ Of the 453 patients eligible for analysis, 430 underwent complete surgical staging that included lymphadenectomy. The median survival was 62.6 months for homologous carcinosarcoma, 22.7 months for heterologous carcinosarcoma, and 20.6 months for LMS. The overall recurrence rate for homologous carcinosarcoma was 56%.

In patients with LMS, lymph vascular space involvement and involvement of the cervix and isthmus were common, whereas lymph node metastases, adnexal metastases, and positive peritoneal cytology were infrequent findings. The only surgicopathologic finding that correlated with progression-free interval was the mitotic index.^{2, 35, 37, 65} While there were no treatment failures among the three women who had less than 10 mitoses per 10 high-power fields, 61% of women with 10–20 mitoses per 10 high-power fields and 79% of women with greater than 20 mitoses per 10 high-power fields developed recurrences.

In contrast to patients with LMS, surgicopathologic factors of carcinosarcoma that related to progression-free interval included adnexal spread, lymph node metastasis, histologic cell type (heterologous vs homologous), and the grade of sarcoma. Of note, patients with carcinosarcoma had high rates of nodal and adnexal metastases and positive peritoneal cytology. Pelvic nodes were involved twice as often as aortic nodes (15% vs 7.8%), and both nodal groups were involved in 5% of the patients.³⁵

Morcellation

In patients undergoing laparoscopic hysterectomy for a presumed uterine sarcoma, power morcellation should not be attempted. Wright (2014) used a large insurance database to identify 36,470 women undergoing minimally invasive hysterectomy with power morcellation performed to demonstrate that the prevalence of uterine cancers was 27 per 10,000 (99 cases).⁶⁶ In addition to this cohort, 26 cases of other gynecologic malignancies were found (a prevalence of 7/10,000), 39 uterine neoplasms of uncertain malignant potential (11/10,000), and 368 cases of endometrial hyperplasia (101/10,000).⁶⁶

LMS

Morcellation of a uterine leiomyosarcoma is associated with a worsened outcome.⁶⁷ Park et al.⁶⁸ reported a series in which the 5-year disease-free survival (DFS) was 40% in women who underwent tumor morcellation and were subsequently diagnosed with leiomyosarcoma, compared to 65% in those with leiomyosarcoma who had not undergone morcellation (p = 0.04). Similarly, the 5-year OS was 46% after morcellation compared to 73% in those not morcellated (p = 0.04).⁶⁸

ESS

In one study, tumor morcellation in women with ESS was associated with a lower 5-year DFS compared with those who did not have a morcellation (55% vs 84%, respectively, odds ratio 4.03, 95% CI 1.06–15.3).⁶⁹ However, no significant impact on OS was reported.

Surgical treatment

The initial therapy for sarcomas of the gynecologic tract is surgical except for embryonal rhabdomyosarcomas. Patients with uterine sarcoma require total abdominal hysterectomy and careful staging including pelvic and paraaortic lymph node sampling. In patients with carcinosarcoma limited to the uterus by pathologic staging, the cytologic presence of malignant cells in the peritoneal washings is a poor prognostic factor.

The ovaries could be retained in premenopausal patients with LMS because this appears to improve their prognosis.^{12, 64, 70, 71} However, a bilateral salpingo-oophorectomy should be performed in all other patients, including those with low-grade ESS,^{72, 73} because these tumors may be hormone dependent or responsive and have a propensity for extension into the parametria, broad ligament, and adnexal structures.

For carcinosarcoma, a high percentage of patients with clinical stage I or II disease is upstaged at the time of laparotomy^{74, 75}; thus, it appears reasonable to surgically stage these patients. There is a paucity of data regarding the role of lymph node sampling in patients with LMS and ESS, but it appears that almost all patients with these sarcomas who have lymph node metastases also have evidence of intraperitoneal disease spread.⁴⁷

Unlike other gynecologic malignancies, there is a role for thoracotomy or video-assisted thoracoscopy in patients with uterine sarcoma metastatic to the lung. Levenback and colleagues reviewed 45 patients whose pulmonary metastases from uterine sarcoma were resected at Memorial Sloan-Kettering Cancer Center, the majority of which were LMS (84%).⁷⁶ The mean survival of patients with unilateral disease (39 months) was significantly greater than that of patients with bilateral disease (27 months). Recurrent or metastatic low-grade ESS may also be amenable to surgical excision of pelvic disease or pulmonary metastases.

Postsurgical therapy for gynecologic sarcomas

Although complete surgical removal is the ideal initial therapy for patients with sarcoma of the gynecologic tract, there is no randomized study proving that surgical cytoreduction influences OS for patients with advanced or recurrent disease. Similarly, the therapeutic benefit of lymphadenectomy has not been proven but is rational. For patients with sarcoma of the uterus or ovary, no formal trial has evaluated the role of lymphadenectomy in addition to hysterectomy and bilateral salpingo-oophorectomy.

For patients with uterine sarcoma, there is no definitive evidence from prospective trials that adjuvant therapy of any type leads to overall improvement in survival. To review the currently understood role of radiotherapy and chemotherapy in sarcomas, LMS is separated from the remaining homologous and heterologous carcinosarcomas because the patterns of relapse for the former are somewhat different from those of the latter group.

Radiation oncology

Radiation therapy for carcinosarcoma

Historically, pre- or postoperative pelvic irradiation has been used as an adjunct to surgery for carcinosarcoma. Many retrospective reviews illustrate this common use.^{60, 75, 80–86} In several reports for carcinosarcoma, the pelvic recurrence rate was 56%, while the distant metastasis rate was 45%. This represents a higher risk of pelvic recurrence than that seen in patients with LMS.^{60, 80–82, 86} It also demonstrates that surgery alone, even for disease apparently confined to the uterus, is inadequate for control of disease in the pelvis. Some but not all studies have shown benefit from postoperative irradiation,^{87, 88} especially in local control.^{60, 84, 88–93}

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Aberrant signaling pathways in cancer Disparities in cancer care Cancer cell immortality: targeting telomerase Tumor suppressor genes Neoplasms of the small intestine, vermiform appendix, and peritoneum and carcinoma of the colon and rectum Infections in patients with cancer

Stay updated, free articles. Join our Telegram channel

Join