Chapter 32: Tumors of the Uterine Corpus

Epidemiology

Endometrial cancer is the most common gynecologic malignancy in the United States. In 2015, the American Cancer Society estimated there were 54,870 new cases and over 10,170 endometrial cancer–related deaths (¹). Approximately 75% of women diagnosed with endometrial cancer are diagnosed at an early stage and have a 5-year overall survival of 74% to 91% (²). For women with stage III or IV disease, reported 5-year overall survival rates are 57% to 66% and 20% to 26%, respectively (²). This disease primarily affects women in their postmenopausal and perimenopausal years, and the average age at diagnosis is 61 years (²).

Risk Factors

The main risk factors for endometrial cancer are age, obesity, diabetes, and exposure to excess estrogen without adequate opposition by progesterone. This includes the historical use of exogenous unopposed estrogen therapy, the use of estrogen agonists (such as tamoxifen), and physiological states that lead to excess endogenous estrogen. Excess endogenous estrogen can be found in women with obesity, chronic anovulation, early age at menarche, nulliparity, late age of menopause, and in the setting of rare estrogen-secreting tumors.

Tamoxifen is a selective estrogen receptor modulator that is used for chemoprevention of breast cancer and as a part of adjuvant therapy for estrogen-receptor–positive breast cancer; it may also be used to treat metastatic breast cancer. It acts as an antagonist on estrogen receptors in the breast but also has a weak agonist effect on the estrogen receptors in the endometrium. These estrogenic effects on the endometrium lead to a significant increase in the risk of developing endometrial cancer, with a relative risk of 2.13 and an absolute annual risk of about 2 per 1,000 patients taking tamoxifen (³). In a study by the National Surgical Adjuvant Breast and Bowel Project (NSABP), women taking tamoxifen (20 mg/d) developed endometrial cancer at a rate of 1.6 per 1,000 patient years, compared to 0.2 per 1,000 patient years among women taking a placebo. At the same time, the 5-year survival rate from breast cancer was 38% higher among women taking tamoxifen compared to women in the placebo group. This suggested that the relatively small risk of developing endometrial cancer was outweighed by the greater survival benefit for women with breast cancer (⁴).

Obesity is a well-established risk factor for the development of endometrial cancer. A meta-analysis of 19 prospective studies showed that for every 5 kg/m² increase in body mass index (BMI), a women’s risk of developing endometrial cancer significantly increases (⁵). One explanation for these findings is that women who are obese have higher levels of endogenous estrogen because of the conversion of androstenedione into estrone and the aromatization of androgens to estradiol, which occurs in the peripheral adipose tissue. There are additional proposed mechanisms for this association, including alterations in insulin signaling and insulin resistance, alterations in expression of circulating adipokines, as well as alterations in pathways involving inflammation (^6,7,8). Obesity increases the risk of endometrial cancer in both post- and premenopausal women. In fact, the majority of young, premenopausal women diagnosed with endometrial cancer have a BMI greater than 30 (⁹).

Lynch syndrome is an autosomal dominant hereditary cancer syndrome characterized by an increased risk in endometrial cancer, colon cancer, and several other malignancies. It is associated with a germline defect in a DNA mismatch repair gene (MutL homolog 1 [MLH1], MutS homolog 2 [MSH2], MutS homolog 6 [MSH6], postmeiotic segregation 2 [PMS2]). Women who have Lynch syndrome have a 15% to 66% lifetime risk of developing endometrial cancer, a 15% to 68% lifetime risk of developing colon cancer, and a 1% to 20% lifetime risk of developing ovarian cancer, depending on the specific mutation involved (^10,11).

Diagnosis and Screening

Screening

Except for patients who have been diagnosed with Lynch syndrome, there is insufficient evidence to support screening for endometrial cancer in asymptomatic women in the general population, even those with risk factors such as tamoxifen use, history of unopposed estrogen, obesity, or diabetes. The American Cancer Society currently recommends that all women be informed about the risks and symptoms of endometrial cancer at menopause. Women should alert their physicians of any unexpected bleeding or spotting in the postmenopausal years and abnormal uterine bleeding in the pre- or perimenopausal years.

The American Cancer Society recommends yearly screening with endometrial biopsy beginning at age 35 for women who are known to carry Lynch syndrome–associated mutations, those who have a family member known to carry this mutation, or women from families with an autosomal dominant predisposition to colon cancer in the absence of genetic testing for Lynch syndrome (¹²). Prophylactic hysterectomy and salpingo-oophorectomy have been shown to be effective in decreasing the risk of endometrial and ovarian cancer and should be offered to women with Lynch syndrome after childbearing is complete (¹³).

Diagnosis

The primary symptom of endometrial cancer is postmenopausal bleeding or abnormal uterine bleeding in pre- or perimenopausal women. A histologic diagnosis is most commonly obtained by office endometrial sampling or dilation and curettage.

Histopathology

Endometrial Hyperplasia

Women with histologic diagnoses of endometrial hyperplasia often present with postmenopausal bleeding or menometrorrhagia. The risk factors associated with the development of endometrial hyperplasia, including obesity and unopposed estrogen, are similar to those associated with endometrial cancer (¹⁴). The World Health Organization defined endometrial hyperplasia based on two characteristics: (1) simple or complex glandular/stromal architectural pattern and (2) the presence or absence of nuclear atypia. The presence of nuclear atypia correlates highest with the risk of developing malignancy. In this spectrum, those with simple hyperplasia without atypia are least likely to develop endometrial carcinoma, while women with complex hyperplasia with atypia are most likely to develop carcinoma. In a long-term follow-up study conducted by Kurman et al, endometrial carcinoma occurred in simple, complex, simple atypical, and complex atypical hyperplasia in 1%, 3%, 8%, and 29% of cases, respectively (¹⁵). A report from one meta-analysis showing a wider range of risk of cancer progression based on four prospective follow-up studies is summarized in Table 32-1 (¹⁶).

Endometrioid Carcinoma

Endometrioid adenocarcinoma is the most common subtype of endometrial cancer, accounting for 75% to 80% of endometrial cancer. Endometrioid adenocarcinomas are graded 1 to 3, based on degree of differentiation. Grade 1 tumors are well differentiated, composed of glands resembling normal endometrial tissue (Fig. 32-1A). In less-differentiated tumors, glandular formations are less evident or are replaced with solid areas (Fig. 32-1B). The percentage of tumor consisting of these nonsquamous or nonmorular solid areas is the criterion for grading, with grade 1 having a 5% or less solid component, grade 2 having more than 5% but less than or equal to 50%, and grade 3 having greater than 50%. Notably, the presence of nuclear atypia that does not correlate with the architectural features should lead to a one-step upgrade of the tumor.

Papillary Serous Carcinoma

Papillary serous carcinomas are highly aggressive tumors that should be distinguished from other types of uterine carcinoma. Papillary serous carcinomas tend to occur in older women, are usually present in advanced stages, and represent 1% to 5% of cases. Deep myometrial invasion, extensive lymphovascular space invasion (LVSI), and extrauterine spread are common. Intraperitoneal involvement is often observed even when the primary lesion is localized. The histologic features of papillary serous carcinoma resemble those of ovarian or tubal carcinoma. Cytologic atypia is so prominent that papillary serous carcinomas should always be graded as poorly differentiated tumors.

Clear Cell Carcinoma

Another aggressive form of uterine carcinoma, clear cell carcinoma, represents 5% to 10% of endometrial cancer cases and should also be considered a high-grade tumor. Clear cell carcinoma tends to occur in older women. Morphologically, it resembles clear cell carcinoma arising from other sites (eg, vagina, cervix, ovary), but clear cell carcinoma of the uterine corpus, unlike that of the vagina or cervix, is not associated with intrauterine exposure to diethylstilbestrol. The microscopic appearance varies and can include solid clear cell features, prominent glycogen content, and a glandular, tubulocystic, or papillary pattern. The cells have abundant clear or eosinophilic cytoplasm, sometimes containing periodic acid–Schiff-positive hyaline globules (Fig. 32-2). The nuclei show marked atypia with frequent mitotic figures. The characteristic “hobnail” appearance is seen as cells with scanty cytoplasm and nuclei protruding into the lumen of the gland. At least half of the clear cell carcinomas are admixed with uterine papillary serous carcinoma (UPSC); this has contributed to ideas that the poor prognosis associated with clear cell carcinoma is due to the presence of UPSC.

Mixed Cell Type

As the name implies, mixed-cell endometrial cancers contain more than one type of tumor. The proportion of the minor component should exceed 10% for a tumor to be designated a mixed cell type.

Carcinosarcomas

Carcinosarcomas (also called mixed müllerian malignant tumors, MMMTs), consist of two components: malignant epithelial and mesenchymal tissue. In homologous MMMT, the sarcomatous component can be tissue that is normally found in the uterus; endometrial stromal sarcoma (ESS) or leiomyosarcoma are common. The so-called heterologous type of MMMT consists of tissue that does not normally appear in the uterus; rhabdomyosarcoma and chondrosarcoma are two common types (^17,18,19). The MMMTs are believed to be aggressive carcinomas, rather than true sarcomas.

Staging

In 1988, the International Federation of Gynecology and Obstetrics (FIGO) committee established a surgical staging system for endometrial cancer (FIGO1988) (Table 32-2). In 2009, this staging system was updated to better stratify patients based on clinically relevant prognostic factors (FIGO 2009, Table 32-3).

Prognostic Factors

There are clinical and pathologic factors that can be used to predict the behavior of endometrial cancer. Clinical factors include age, ethnicity, and stage of disease. Pathologic factors include histologic type, grade, depth of myometrial invasion, LVSI, and spread of tumor outside the uterine cavity, including the retroperitoneal lymph nodes.

Clinical Factors

Studies have shown that outcomes for women with endometrial cancer are more favorable among younger patients. A retrospective study showed that patients less than 45 years old were statistically more likely to have endometrioid histology, grade I tumors, and stage IA disease, while women over age 65 were significantly more likely to have papillary serous histology and grade 3 tumors. A subset analysis of patients greater than 75 years of age showed an increase in the percentage of patients with aggressive papillary serous histology, higher-grade (grade 3) disease, and advanced-stage disease compared to those less than 45 years old. Evaluation of patients with endometrioid tumors revealed a similar pattern of deeper myometrial invasion and higher tumor grade as age increased (²⁰). Other studies have shown that the risks of locoregional relapse and death among patients older than 60 years were twice as high as the risks among those aged 60 years or younger (²¹).

Studies in the United States have demonstrated that white women are more likely to be diagnosed with endometrial cancer than black women. Interestingly, however, black women diagnosed with endometrial cancer are more likely to die of their disease. They tend to be diagnosed at a later stage, with higher-grade tumors, and are more likely to have papillary serous histology. However, after controlling for these factors as well as other comorbid factors, a higher disease-specific mortality rate among black women still exists (²²).

Pathologic Factors

With regard to histologic subtype, papillary serous carcinoma and clear cell carcinoma are associated with worse prognosis than endometrioid carcinoma. Increasing tumor grade is associated with increased myometrial invasion, pelvic and para-aortic node involvement, recurrence, and a poorer overall survival (²³) (Table 32-4, Fig. 32-3).

Lymphovascular space invasion, an independent risk factor for recurrence, is present in about 15% of endometrial cancer cases. Multivariate analyses have shown angioinvasion to be associated with survival in endometrial cancer regardless of disease stage. Independent of tumor grade or depth of myometrial invasion, LVSI is also associated with risk of pelvic and para-aortic node involvement and thus should lead to surgical nodal assessment or empiric radiation in patients with unstaged disease (²³).

The depth of myometrial invasion is associated with degree of tumor differentiation, LVSI, lymph node involvement, extrauterine spread, recurrence, and overall survival (Fig. 32-4, Table 32-5). Adnexal metastasis correlates strongly with metastatic involvement of pelvic and para-aortic nodes; in one study, 32% of patients with adnexal involvement had positive pelvic nodes, as compared with 8% of those without adnexal disease (²³). However, the alternative possibility of simultaneous ovarian and endometrial primaries must be ruled out as the surgical management and prognosis are vastly different. Up to 11% of patients with clinical stage I or II disease have lymph node involvement (²³). The risk of para-aortic node involvement increases if the pelvic nodes are positive for tumor.

The current FIGO staging system does not incorporate peritoneal washings as the evidence for their prognostic value is weak. Positive findings on peritoneal cytology are often associated with other unfavorable prognostic factors, such as deep myometrial invasion, cervical extension, and extrauterine spread. In particular, UPSC demonstrates a high rate of positive washings. Even in clinical stage I disease that is confined to the uterus, cytologic findings have been positive in as many as 16% to 17% of cases (²⁴). However, in at least two studies, the 5-year survival rates for patients with cytologically positive disease but no other risk factors have been higher than 90% (²⁴) (Fig. 32-5).

In conclusion, these prognostic factors correlate with clinical course of disease, recurrence, and risk of death. However, the relative importance of each factor is not always clear because of interrelations and interdependence among them. Results of a multivariate analysis by Zaino and colleagues of factors affecting survival in stages I and II endometrial cancer are summarized in Table 32-6 (²⁵).

Evaluation

When evaluating a patient diagnosed with endometrial cancer, in addition to a history and physical examination, a chest x-ray should be performed, and imaging of the abdomen and pelvis should be considered. The measurement of cancer antigen 125 (CA125) may occasionally be useful (Fig. 32-6).

At MD Anderson, for patients with disease confined to the uterus, we recommend primary surgery with total hysterectomy and bilateral salpingo-oophorectomy, except in cases with significant medical comorbidities (see Fig. 32-6). Lymph node evaluation is based on intraoperative evaluation of grade and depth of invasion. Patients with gross cervical involvement may either undergo radical hysterectomy, bilateral salpingo-oophorectomy, with full staging, including pelvic and para-aortic lymph node sampling, or they may undergo primary radiation therapy (see Fig. 32-6). For disease that is not confined to the uterus, surgical debulking is considered (see Fig. 32-6).

Surgery

The standard staging procedure for endometrial cancer includes total hysterectomy, bilateral salpingo-oophorectomy, and pelvic and para-aortic lymphadenectomy. However, the indications for performing a pelvic and para-aortic lymphadenectomy for staging remain controversial among gynecologic oncologists. Some surgeons practice routine pelvic or pelvic and para-aortic lymphadenectomy for all patients with endometrial cancer, whereas others perform these procedures only for those patients with an increased risk of lymph node metastases based on preoperative and intraoperative assessments. The decision to selectively perform lymph node staging is based on multiple concerns. First, many believe there has been insufficient or unconvincing evidence in the literature that routinely performing lymphadenectomy for all patients results in improved overall survival. Second, performing a lymphadenectomy is associated with additional risks for the patient, including the risk of additional surgical complications and the long-term development of lower extremity lymphedema. Finally, there is a very low rate of lymph node metastases in patients with early disease, specifically those with low-grade endometrioid adenocarcinoma with less than 50% invasion of the myometrium. The absence of LVSI and a tumor size of 2 cm or less in diameter also add reassurance that the lymph nodes are not involved (²⁶)

Because of this, intraoperative gross inspection and frozen section of the uterus are commonly performed to evaluate the extent of uterine disease (see Fig. 32-6). This information may be used to further guide whether a lymph node dissection will be performed. The preoperative histologic findings and intraoperative pathologic findings that correlate with an increased risk of nodal involvement include grade 3 endometrioid histology; other aggressive histologies, including clear cell, serous, or squamous carcinoma; tumor invasion of more than half of the myometrium; tumor size greater than 2 cm; and the presence of extrauterine disease (²³). In the presence of these variables, there is a benefit for surgical resection of the pelvic and para-aortic lymph nodes (see Fig. 32-6).

In the process of lymph node evaluation, mere palpation of the fatty tissue of node-bearing areas is inadequate for the evaluation of nodal status because only about half of metastatic nodes are enlarged, and fewer than 30% can be identified as abnormal on palpation (²³). A representative lymph node sampling or lymph node dissection should be performed to adequately evaluate the nodal basins.

The use of sentinel lymph node biopsy to identify patients with lymph node metastases without performing a complete lymph node dissection is currently under study. Theoretically, this could be performed even in patients at low risk for metastases, thereby avoiding the additional risks associated with performing a lymphadenectomy while accurately identifying those in need of additional adjuvant therapy.

Minimally Invasive Approaches

In two reports by the Gynecologic Oncology Group (GOG), laparoscopic surgical methods were compared to conventional laparotomy. The GOG LAP-2 study is the largest randomized trial ever performed in endometrial cancer. In this phase III prospective, multi-institutional, randomized study, 2,616 patients with stage I to IIA uterine cancer were assigned to laparoscopy (n = 1,696) or laparotomy (n = 920). Study end points were morbidity and mortality at 6 weeks, length of hospitalization, failure to complete laparoscopy, site of recurrence, and recurrence-free survival. Consistent with previous studies, laparoscopy required longer operative time but led to fewer postoperative adverse events and shorter hospitalization. The rate of intraoperative complications was similar. One unexpected outcome was that as many as 26% of patients in the laparoscopy arm were converted to laparotomy. Laparoscopic failure was associated with increasing age and body mass index (²⁷).

In a companion report, patients undergoing surgical staging via laparoscopy versus laparotomy were assessed with quality-of-life (QOL) measures. The study’s objective was to compare the QOL of patients with endometrial cancer undergoing surgical staging via laparoscopy versus laparotomy. Although patients in the laparoscopy arm had overall better QOL measures at 6 weeks postsurgery, except for better body image in the laparoscopy arm, no difference was detected between the two arms at 6 months (²⁸).

Robot-assisted surgery is a newer minimally invasive technique commonly offered to women with endometrial cancer. A recent systematic review included eight studies, with a total of 589 patients undergoing surgery for endometrial cancer. When robot-assisted surgery was compared to laparotomy, it was found to be associated with lower estimated blood loss at time of surgery, lower rates of wound complications, and lower rates of other postoperative complications (including stroke, ileus, nerve palsy, lymphedema). The duration of surgery was significantly longer in robot-assisted surgery compared to laparotomy (²⁹).

The decision to perform open or minimally invasive surgery depends on patient preference, the skill and experience of the surgeon, the availability of the equipment, the size of the uterus, the parity of the patient, and the patient’s medical condition. At MD Anderson, minimally invasive approaches, both laparoscopic and robotic, are becoming the standard of care for the majority of women with endometrial cancer.

Adjuvant Therapy

At MD Anderson, we stratify patients according to features of the disease and various prognostic factors (Fig. 32-7). The risk group helps guide the choice of adjuvant treatment. Women with grade 1 or 2 endometrioid tumors with no myometrial invasion or LVSI are considered to have low-risk disease. Low-risk disease is associated with an excellent prognosis, with 5-year survival approaching 90% and a low risk for recurrence after surgery. Therefore, in general, we do not recommend adjuvant therapy for this group of patients (see Fig. 32-7). Women with stage III disease, regardless of histology or grade, and women with serous carcinoma or clear cell carcinoma at any stage are considered to have high-risk disease and typically undergo adjuvant therapy. Intermediate-risk disease, therefore, includes women with endometrial cancer confined to the uterus that invades the myometrium or the cervix. The intermediate-risk group can be further stratified based on the presence of specific prognostic factors, including outer one-third myometrial invasion, grade 2 or 3 disease, and the presence of LVSI. Women are considered to have high-intermediate-risk disease if they are (1) any age with all three factors, (2) age 50 to 69 with two factors, and (3) age greater than or equal to 70 years old with one factor. For patients with intermediate-risk disease, there are ongoing controversies regarding the benefit of adjuvant therapy.

Low-Risk Disease

As mentioned, most patients with low-risk disease are unlikely to benefit from adjuvant therapy. Therefore, it is not generally recommended for this group (see Fig. 32-7).

Intermediate-Risk Disease

A number of trials have addressed the question of adjuvant therapy in intermediate-risk endometrial cancers. At this time, there is no evidence that adjuvant therapy improves overall survival for women with intermediate risk disease. For intermediate-risk patients, adjuvant radiation therapy reduces the incidence of pelvic recurrence without prolonging overall survival (Fig. 32-8) (³⁰). Vaginal irradiation alone is adequate for preventing pelvic recurrence in patients who are at risk of only isolated vaginal recurrence. At MD Anderson, vaginal irradiation alone is offered to patients with low-intermediate risk (see Fig. 32-7). Vaginal irradiation has a higher therapeutic ratio than whole-pelvis radiation because it produces fewer long-term sequelae. The rate of recurrence after vaginal irradiation is low in women with negative or unknown lymph node status (^31,32). Patients with high-intermediate-risk disease at MD Anderson are offered adjuvant external pelvic radiation with or without brachytherapy (see Fig. 32-7).

Two large, randomized controlled trials compared external beam radiotherapy following surgery to no adjuvant therapy in intermediate-risk endometrial cancer. For GOG-99, 392 surgically staged patients were randomized to external beam radiation (50.4 Gy) or no treatment (control) (³¹). The study included patients with any degree of myometrial invasion, adenocarcinoma of any grade, with FIGO 1988 stages IB, IC, and occult stage II disease. The 2-year pelvic recurrence rate was 3% in the radiation arm versus 12% in the control arm (relative hazard 0.42; P = .007). In subgroup analysis, the treatment difference was particularly evident among the high-intermediate-risk subgroup (2-year cumulative incidence rate 6% vs 26% in radiation vs no radiation arms; relative hazard ratio [HR] 0.42). In fact, it was this study that defined the high-intermediate-risk subgroup described previously. In this study, the high-intermediate-risk subgroup was defined as those with: (1) moderate to poorly differentiated tumor, presence of lymphovascular invasion, and outer third myometrial invasion; (2) age 50 or greater with any two risk factors listed; or (3) age of at least 70 with any risk factor listed. However, the estimated 4-year survival was 92% in the radiation arm versus 92% for the control arm, which was not significantly different (HR 0.86; P = .557). The authors concluded that adjunctive radiation in early-stage intermediate-risk endometrial carcinoma decreases the risk of recurrence but should be limited to patients whose risk factors fit a high-intermediate-risk definition.

The PORTEC-1 (Post-Operative Radiation Therapy in Endometrial Carcinoma) study randomized 714 women with intermediate-risk endometrial cancer following surgical staging to either external beam radiotherapy (46 Gy) or no adjuvant treatment (³⁰). At 5 years, local recurrence was significantly lower for patients in the experimental arm (4%) versus the control arm (14%) (³⁰). At 10-year follow-up, the local recurrence rates were similar: 5% in the radiotherapy arm versus 14% in the control arm (³³). There was no significant difference in distant recurrence rates at either 5 or 10 years and no overall survival benefit seen at either time point.

In both GOG-99 and PORTEC-1, patients with initial uterine-confined disease had local recurrences in the vagina, prompting evaluation of vaginal brachytherapy as a possible adjuvant therapy. Previously, Piver et al conducted a randomized controlled trial comparing (1) hysterectomy alone, (2) preoperative uterine radium followed by hysterectomy, and (3) hysterectomy and postoperative vaginal radium. This study reported no difference in local control rates, disease-free survival, or overall survival at a follow-up of 10 years (³⁴).

More recently, PORTEC-2 randomized 427 women with high-intermediate-risk disease to vaginal brachytherapy or pelvic radiation. At 45 months, there were no statistically significant differences between the two treatment modalities in terms of locoregional recurrence, distant metastases, or 5-year disease-free survival. Vaginal brachytherapy was associated with a significantly lower rate of side effects, including treatment-related diarrhea and other bowel symptoms (^35,36).

At this time, the use of adjuvant chemotherapy to treat intermediate-risk disease is not supported by clinical evidence. Interestingly, however, it has been proposed that chemotherapy with or without vaginal cuff radiation may provide superior results to pelvic radiation in intermediate-risk patients. There are ongoing randomized clinical trials, including GOG 249 and PORTEC-3, attempting to evaluate this combined modality therapy in the high-intermediate-risk patient population.

High-Risk Disease

Women with high-risk disease have either serous or clear cell adenocarcinoma (any stage) or have pathologic stage III disease with extrauterine involvement. There are fewer high-quality studies to guide adjuvant therapy in this group. Treatment may involve radiation, chemotherapy, or a combination of the two modalities, based on histology, stage, and other factors (see Fig. 32-7). For patients with nodal disease, at MD Anderson Cancer Center (MDACC), we typically treat with chemoradiation followed by chemotherapy with four cycles of carboplatin and paclitaxel. A recent meta-analysis evaluated the use of chemotherapy alone versus radiation alone versus chemoradiation in women with stage III or IV endometrial cancer. Compared with radiation, the use of a combination platinum-based chemotherapy regimen resulted in a statistically significant improvement of overall survival (HR 0.75; 95% CI, 0.57-0.99) and progression-free survival (HR = 0.74; 95% CI, 0.59-0.92) (³⁷). The MD Anderson algorithm further delineates our practice pattern for these patients (see Fig. 32-7).

Recurrent or Metastatic Disease

Women with disease outside the pelvis, either at their initial diagnosis or at the time of recurrence, are considered to have metastatic endometrial cancer. In general, these patients carry a poor prognosis and may benefit from systemic chemotherapy.

The GOG 209 study compared carboplatin plus paclitaxel to the combination of paclitaxel, doxorubicin, and cisplatin (TAP) in 1,300 women with chemotherapy naïve stage III, IV or recurrent metastatic endometrial cancer. Both regimens were administered every 3 weeks for a total of seven cycles. The data are not yet mature; however, initial findings were presented at the 2012 Society of Gynecologic Oncology Annual Meeting. Compared to TAP, carboplatin plus paclitaxel had a similar overall response rate (51% in each arm), similar progression-free survival (median of 13 months in each arm), and similar overall survival (37 months for carboplatin and paclitaxel vs 40 months for TAP). Importantly, the toxicity in the carboplatin-and-paclitaxel arm was significantly lower than the TAP arm. Peripheral neurotoxicity was 19% in the carboplatin/paclitaxel arm compared to 26% in patients assigned to receive TAP. Based on this study, carboplatin plus paclitaxel may often be preferred.

There is ongoing research about the use of biologic therapies in this setting. Single-agent bevacizumab demonstrated clinical activity and was well tolerated in GOG 229E (³⁸). It is currently being studied in combination with cytotoxic chemotherapeutic regimens in GOG 86P, a recently closed randomized trial. The use of mTOR inhibitors has also been under investigation. A recent phase II trial of the mTOR (mechanistic target of rapamycin) inhibitor everolimus plus letrozole demonstrated a high clinical benefit rate and response rate (³⁹). Additional research with these agents is also under way.

Endocrine Therapy

Progestins have been used to treat advanced endometrial cancer for over 50 years. These medications tend to be well tolerated with relatively minor side effects, including weight gain, thrombophlebitis, headache, and occasional hypertension. Response rates tend to be modest, ranging from 11% to 24% (^40,41). In GOG 121, women with no previous chemotherapy or hormonal treatment were given megesterol acetate (800 mg/d). The overall response rate was 24%, with progression-free survival of 2.5 months and overall survival of 7.6 months (⁴⁰). Unlike in breast cancer, there does not appear to be a dose-response effect for progestins in endometrial cancer. In a randomized trial of oral medroxyprogesterone acetate (GOG 81), women who received a low-dose regimen (200 mg/d) actually had a higher response rate compared to women who received the high-dose regimen (1,000 mg/d) (⁴¹).

Certain characteristics improve the likelihood that a patient will have a favorable response to hormone therapy. These include having a low tumor grade (1 or 2), endometrioid histology, the presence of estrogen and progesterone receptors, having a longer disease-free interval, and being asymptomatic or minimally symptomatic.

Tamoxifen has been shown to have some efficacy in this population, with improved rates among women with hormone receptor–positive disease. Interestingly, it has been shown that short-term use of tamoxifen may lead to an increase in progesterone receptors. In GOG 153, women were given a regimen of tamoxifen for 3 weeks, alternating with megestrol acetate for 3 weeks. The overall response rate was 27%, with a median progression-free survival of 2.7 months and overall survival of 14 months (⁴²). Letrozole and anastrazole are both aromatase inhibitors that have been evaluated in advanced or recurrent endometrial cancer in phase II trials. Both agents have shown a response rate of less than 10% (^43,44).

Surgery for Relapsed Disease

At MD Anderson, we evaluate all patients who have experienced an isolated central pelvic recurrence after radiotherapy for radical pelvic surgery or pelvic exenteration; these procedures remain the only potentially curative options that offer the possibility of long-term survival. However, given the high incidence of major postoperative complications, such as urinary/intestinal tract fistulas, pelvic abscesses, septicemia, pulmonary emboli, and cerebrovascular accidents, we consider only selected patients with isolated central recurrences for this treatment.

Postoperative Surveillance

There are no prospective studies to guide frequency of postoperative follow-up. At MD Anderson, for patients with low-risk disease, we schedule office visits every 6 months for the first year and annually during years 2 to 5. For patients with intermediate- and high-risk disease, we schedule office visits every 3 months for year 1, every 4 months for year 2, and every 6 months for years 3 to 5 (Fig. 32-9). At each visit, patients undergo a physical and pelvic examination. Performing routine vaginal cytology in endometrial cancer survivors is not recommended by the SGO. This is because multiple studies have suggested that Papanicolaou tests are not as effective in detecting endometrial cancer recurrences compared to physical examination alone (^45,46) and are costly and inefficient (⁴⁷). In one of these studies, abnormal vaginal cytology had a sensitivity of 40% and specificity of 88% for detecting a vaginal recurrence. The positive predictive value was 7.3%, and negative predictive value was 98.4% (⁴⁶).

Serum CA125 level measurements may be helpful for certain patients. In patients with elevated serum CA125 levels at diagnosis, they can be used as a marker for recurrent disease. In one study, among patients with advanced endometrial cancer who initially had high CA125 levels, 26% to 58% showed elevated CA125 levels at recurrence (⁴⁸). The use of serial CA125 assays is most beneficial in diagnosing recurrence in a high-risk population, although there have been no studies to show that their use improves survival. False elevations may occur following radiation therapy.

Hormone Replacement Therapy Following Treatment

The use of estrogen replacement therapy in women who have been treated for endometrial cancer is controversial. Most of the evidence regarding the effects of hormone replacement therapy in women with previously treated endometrial cancer comes from retrospective studies. In one study, Chapman et al found no significant difference in disease-free survival rates between women who used estrogen replacement therapy and those who did not (⁴⁹). However, in this study, the group that used estrogen had earlier-stage disease with less depth of invasion compared to the control group. A subsequent cohort study by Suriano et al, who matched cases for tumor status and treatment, also found lower recurrence rates and longer disease-free survival among women who used estrogen (⁵⁰). Notably, 49% of patients in the estrogen replacement group in that study were also given progestin, which might have influenced the apparent improvement in outcome.

In a study reported in 2006, the GOG conducted a double-blind, randomized, phase III trial to evaluate the safety of estrogen replacement therapy following treatment for endometrial cancer (GOG 137) (⁵¹). Within 20 weeks following surgery, 1,236 women with a history of stage I or occult stage II endometrial cancer were enrolled. All enrolled patients had an indication for treatment with estrogen replacement therapy, including hot flashes, vaginal atrophy, increased risk of cardiovascular disease, or an increased risk of osteoporosis, and were randomized to receive estrogen replacement or placebo for 3 years following surgery. This study closed early when the results of the Women’s Health Initiative (WHI) study showed increased overall risks in the estrogen and progestin arm (⁵²). Based on the results of the WHI study, GOG 137 closed prematurely, and the study was left with insufficient power to detect a difference in patients with intermediate- or high-risk early endometrial cancers. The results of this study are inconclusive. However, the authors noted that in the low-risk population, the absolute recurrence rate (2.1%) was low. The relative risk of recurrence/death in the estrogen replacement therapy group was 1.27 compared with the placebo group (80% CI, 0.916-1.77).

At MD Anderson, we approach the use of estrogen replacement therapy on an individual basis. Overall risk of disease recurrence in low-risk patients is low, even in the setting of exogenous estrogen replacement. Therefore, estrogen replacement therapy may be considered for this population. If adjuvant therapy is given following surgery, we recommend a 6- to 12-month waiting period prior to initiation of hormone replacement therapy.

Carcinosarcomas

Carcinosarcomas, or MMMTs, account for less than 5% of malignant neoplasms of the uterus (^{53,54,55,56,57}). The gross appearance of a polypoid carcinosarcoma mass protruding into the endometrial cavity is shown in Fig. 32-10. Although not pathognomonic of carcinosarcoma, this feature in a postmenopausal woman should alert the clinician to the possibility of carcinosarcoma.

Carcinosarcomas, although historically classified as a sarcoma, are now thought to represent a dedifferentiated variety of endometrial carcinoma (⁵⁸). Epidemiologically, carcinosarcomas usually appear in older women aged 60 or above (^53,54,59).

Treatment

Surgery

Carcinosarcoma is an aggressive type of endometrial carcinoma, and surgical management of carcinosarcoma includes total abdominal hysterectomy with bilateral salpingo-oophorectomy and pelvic and aortic lymph node dissection (^60,61,62,63). Peritoneal cytology, omentectomy, and peritoneal biopsies are also recommended (^{60,64,65,66,67}). Approximately 20% to 60% of patients will have more advanced disease during surgical staging (^55,68,69). A recent study reported that in patients with stages I to III uterine carcinosarcomas, 5-year overall survival, disease-free survival, and median survival were significantly improved for patients receiving lymph node dissection compared to those who did not (⁷⁰). In regard to the role of cytoreductive surgery in carcinosarcoma, a recent series of 44 patients demonstrated increased overall survival in those patients who had a complete cytoreductive surgery versus those that had residual disease (52.3 vs 8.6 months, respectively) (⁷¹). Complete cytoreductive surgery should be attempted, when possible, in patients with carcinosarcoma.

Adjuvant Therapy

Adjuvant therapy for stage IA tumors is left to the discretion of the provider as studies have been inconclusive, and the benefit of adjuvant therapy in these patients is unclear (^72,73). A phase III study by the European Organization for Research and Treatment of Cancer of radiotherapy versus observation in early-stage uterine sarcomas showed a trend toward improved local control but higher rates of distant recurrence in carcinosarcomas (⁷³). There was also no significant difference seen in overall survival (⁷³). Another study of adjuvant therapy in early-stage carcinosarcomas showed treatment with chemotherapy to significantly increase progression-free survival; however, overall survival was not significantly increased (⁷²).

For stage IB and greater carcinosarcomas, adjuvant chemotherapy is generally recommended. A meta-analysis showed improved overall survival (HR 0.75; 95% CI 0.60-0.94) and reduced disease progression (HR 0.72; 95% CI, 0.58-0.90) with combination chemotherapy versus single-agent ifosfamide (^74,75,76). Combination regimens included ifosfamide plus paclitaxel or cisplatin (⁷⁶). A phase II study of combination carboplatin and paclitaxel also showed an overall response rate of 54% and median progression-free survival of 7.6 months (⁷⁷). A phase III study, GOG 261, is currently being conducted to further evaluate combination chemotherapy. This study compares combination ifosfamide and paclitaxel to carboplatin and paclitaxel in patients with stage I through IV previously untreated carcinosarcomas (NCT00954174). Examples of these combination chemotherapy regimens are shown in Table 32-7.

Uterine sarcoma is uncommon, accounting for only 3% to 9% of all tumors arising from the uterine corpus (^55,58,68). Most uterine sarcomas originate from mesodermal tissue, although some are derived from specialized müllerian mesenchyme, such as endometrial stroma, and a few originate from nonspecific or nonmüllerian mesenchyme (eg, smooth or skeletal muscle, vessels, or lymphoid tissue). The three most common types of uterine sarcoma are leiomyosarcoma, endometrial stromal sarcoma (ESS), and adenosarcomas. Overall, uterine sarcomas are aggressive tumors with a 5-year overall survival rate of 17.5% to 54.7% (⁷⁸).

The 2009 FIGO staging system includes three classifications for staging uterine sarcomas (Table 32-8) (⁷⁹).

EPIDEMIOLOGY

Leiomyosarcoma and ESS usually affect women in their early 50s. Adenosarcomas have been reported in women 14 to 84 years old; the median age at appearance is in the late 50s (^80,81,82).

Although uterine sarcoma differs somewhat from endometrial cancer in its clinical and pathologic characteristics, some risk factors—such as hypertension, diabetes mellitus, and obesity—are common to both. Previous history of pelvic irradiation has been associated with carcinosarcomas and undifferentiated sarcomas, occurring in 2% to 29% of patients so exposed at an interval ranging from 1 to almost 40 years (^83,84). Prior exposure to pelvic radiation, however, is not felt to be associated with leiomyosarcomas (⁸⁴). Although the link between estrogen and endometrial cancer is well established, no such clear relationship is evident for estrogen and uterine sarcoma. An association between tamoxifen use and uterine sarcoma was confirmed in a study by Bergman et al, showing increased numbers of uterine sarcomas among women taking tamoxifen and sarcomas constituting about 10% of total malignancies in these cases (⁸⁵).

Leiomyosarcoma

Histology

Leiomyosarcomas represent only 1% to 2% of uterine malignancies; however, they are the most frequent of the uterine sarcomas (^{55,56,57,58,86,87}). Gross features of leiomyosarcoma (Fig. 32-11) are typified by variegation at the cut surface (unlike the whorllike surface typical of a benign leiomyoma).

Cells are spindle shaped and arranged in fascicles with eosinophilic cytoplasm. The nuclei are usually elongated with rounded ends, appearing hyperchromatic with coarse chromatin and prominent nucleoli. The most important criterion for the diagnosis of leiomyosarcoma is a high mitotic rate, generally exceeding 15 mitotic figures per 10 high-power fields. Other features, such as extrauterine extension, large size, and infiltrating border, necrosis, and atypical mitotic figures can also aid in the diagnosis (⁵⁸).

Other variants of leiomyosarcoma include myxoid and epithelioid leiomyosarcoma. In myxoid leiomyosarcoma, tumor cellularity is low, cells are separated by myxoid material, and the tumor cells are still characterized by nuclei atypia and high mitotic index (^58,88,89). The tumor cells in epithelioid leiomyosarcoma, in contrast, exhibit epithelioid features distinct from the usual characteristics of leiomyosarcoma (⁹⁰).

Smooth Muscle Tumor of Uncertain Malignant Potential

Uterine smooth muscle tumors with histologic features of necrosis, nuclear atypia, or mitoses that do not meet all diagnostic criteria for leiomyosarcoma fall into the category of smooth muscle tumor of uncertain malignant potential (STUMP) (⁹¹). The STUMPs are a group of smooth muscle tumors for which a diagnosis of sarcoma cannot be made because of their uncertain malignant characteristics. Most of these tumors are associated with favorable prognosis, and only follow-up of the patients is recommended (⁹²). In a study conducted at MD Anderson of 41 cases of STUMP, the recurrence rate was 7%. One of the three recurrences was a leiomyosarcoma; the others were STUMPs (⁹³).

Clinical Presentation

Vaginal bleeding is the most common presenting symptom of uterine sarcomas. Other symptoms, such as back pain, urinary retention or hematuria, gastrointestinal symptoms, or weight loss may indicate an invading or metastasizing tumor (^94,95,96).

Diagnosis

Definitive diagnosis of uterine sarcomas is usually made following hysterectomy, with the diagnosis based on histologic examination. Methods utilizing pulsed Doppler ultrasonography to preoperatively distinguish between a benign leiomyoma and uterine sarcoma are reported, but not clinically useful (⁹⁷). Although useful for detecting extrauterine disease, imaging studies such as computed tomography and magnetic resonance imaging are not specific for diagnosing sarcomas (⁹⁸).

Prognosis and Clinical Course

Many prognostic factors affect outcome in uterine sarcoma, but the most important is the clinical stage at presentation (^{56,87,99,100,101,102}). Disease stage directly affects disease-free and overall survival rates (^56,102,103).

Leiomyosarcomas are aggressive tumors with a recurrence rate of 50% to 70% even at early stages (⁶⁸). The overall 5-year survival rate ranges from 15% to 25%, and the 5-year survival rate for patients with stages I and II is 40% to 70% (^104,105). There has been no consensus on prognostic indicators for leiomyosarcoma. Previous studies have reported mitotic index, cellular atypia, vascular invasion, and tumor size to correlate with survival.

Treatment

Surgery

Surgery remains the mainstay of treatment for uterine sarcoma, regardless of histologic type. However, there is some controversy regarding the need for oophorectomy or lymph node dissection for each subtype of uterine sarcoma. In general, hysterectomy, including oophorectomy with or without lymph node dissection, is the treatment of choice. In addition, morcellation should not be utilized in women undergoing surgery for uterine sarcoma or suspected uterine sarcoma.

Oophorectomy for early-stage leiomyosarcomas, particularly in premenopausal women, is controversial, as the incidence of adnexal metastases is low (⁶⁸). In one study, multivariate analysis showed that oophorectomy for leiomyosarcoma was associated with significantly worse disease-specific survival. Furthermore, case-control investigations implied that ovarian preservation does not adversely affect survival (¹⁰⁶). In another study, ovarian metastases were found in about 10% of all cases, despite most having been detected at an early stage (¹⁰⁷). Hence, the choice of oophorectomy for the treatment of early-stage leiomyosarcoma should be judged on an individual basis.

The need for lymph node dissection is also controversial. Leiomyosarcoma has substantially lower risk of nodal involvement than clinical stage I or stage II endometrial cancer with other risk factors. A GOG study of early-stage sarcoma found that nodal metastases were present in only 3.5% of leiomyosarcoma cases, and the pelvic nodes were twice as likely to be involved as the para-aortic nodes (⁹⁴). However, the patients in this series underwent lymph node sampling rather than intended dissection, so the reported incidence may be falsely low. Goff et al found lymph node involvement in about 27% of leiomyosarcoma cases, but only in those cases that involved recurrent or disseminated intraperitoneal disease (⁹⁴). Two more recent studies identified lymph node metastases in 7% to 11% of patients who underwent lymph node dissection (^106,107). In the series by Kapp et al, the 5-year disease-specific survival rate was 26% in patients who had positive lymph nodes compared with 64% in those who had negative nodes (¹⁰⁷).

Adjuvant Therapy

For early-stage leiomyosarcomas, there has been no clear benefit shown for adjuvant chemotherapy (^108,109,110). In these patients, surveillance versus referral to a clinical trial are reasonable treatment options. Currently, the GOG is conducting a study of stage I completely resected leiomyosarcomas comparing adjuvant chemotherapy to observation (GOG 277). The chemotherapy arm consists of gemcitabine on days 1 and 8 and docetaxel on day 8 repeated every 21 days for up to four cycles (NCT01533207).

Advanced-stage leiomyosarcomas, on the other hand, have a high risk of recurrence, and individuals so affected are typically offered adjuvant therapy. We work closely with our sarcoma medical oncology colleagues for treatment recommendations. Combination chemotherapy, specifically, for leiomyosarcoma improves response rates. Various combinations have been investigated, with response rates ranging from 18% to 53% (^{109,111,112,113,114}). The two most common combination regimens used at MD Anderson are doxorubicin plus ifosfamide or gemcitabine plus paclitaxel. Examples of these combination chemotherapies are shown in Table 32-9.

ENDOMETRIAL STROMAL SARCOMA

Histology

Endometrial stromal sarcoma accounts for about 23% of all uterine sarcomas (^{53,54,55,56,57}). The ESSs originate from endometrial stromal cells that invade the myometrium. Low- and high-grade variants differ in their clinical behavior. Gross features of a low-grade ESS presenting as multiple, small, lobulated tumor masses in the myometrium are shown in Fig. 32-12A.

In fresh specimens, tumor invading the lymphovascular spaces can sometimes be compressed out of the vascular lumen, giving a vermiform appearance. Tumor cells resemble endometrial stromal cells during the proliferative phase of the menstrual cycle and are monotonous and of uniform shape and size (⁹³). The presence of spiral arteriole-like vessels is a characteristic finding of ESS, as is the propensity of the tumor cells to invade the lymphovascular spaces (Fig. 32-12B).

Endometrial stromal sarcoma tumors contain both estrogen and progesterone receptors (⁵⁸). In addition, these tumors are generally positive for CD10 and frequently have a t(1;17) chromosomal translocation that expresses a fusion protein composed of JAZF1 and JJAZ1 (⁵⁸).

Clinical Presentation

Like leiomyosarcomas, vaginal bleeding is the most common presenting symptom. Another common presenting symptom in ESS and adenosarcoma is the appearance of a prolapsed polypoid mass through the cervical os (^94,95,96).

Prognosis and Clinical Course

Low-grade ESSs are indolent tumors (¹¹⁵). Generally, patients with ESS have favorable prognoses; however, recurrence can appear late after primary diagnosis and treatment, even in stage I disease (^116,117). The FIGO stage, depth of myometrial invasion, tumor grade, positive margins, and patient characteristics such as age, race, and menopausal status have been reported to be prognostic factors (^{53,118,119,120}). Mitotic activity and cytologic atypia were found to be important by some (^53,104,121) but not by others (^122,123). Extrauterine and nodal disease is prevalent in ESS.

Treatment

Surgery

Initial treatment for ESS is largely surgical, and hysterectomy with bilateral salpingo-oophorectomy is recommended. Endometrial stromal sarcoma is often sensitive to hormones, and patients with intact ovaries may be at a higher risk for recurrence (^{124,125,126,127}). However, there is no consensus on the absolute need for oophorectomy in early-stage ESS. In a study by Li et al, bilateral salpingo-oophorectomy did not appear to affect time to recurrence or overall survival in patients with stage I ESS (¹²⁵). In a recent report of 384 women with low-grade ESS, lymph node metastasis and ovarian preservation were not significant prognostic factors for survival in this study. Lymph node metastasis was found in 7% of patients (¹²⁶). In two smaller series, only two of the nine patients with ESS who had undergone lymph node sampling were found to have lymph node metastases (^94,121). The choice of lymph node dissection for ESS should be individualized.

Adjuvant Therapy

Observation is generally recommended for stage I ESS. More advanced stages are typically offered additional therapy after surgical resection, although the data are limited given this rare tumor type. As these tumors typically express estrogen and progesterone receptors, hormonal therapy is considered first-line treatment. Limited retrospective data have reported response of metastatic and recurrent ESS to hormonal agents such as medroxyprogesterone acetate or letrazole (^124,128). Tamoxifen is contraindicated in the treatment of ESS due to its stimulatory effect on the endometrial stromal cells (¹²⁸).

Because of its rarity and because low-grade ESS usually responds well to hormonal therapy, ESS has not been well studied in chemotherapeutic trials. A cisplatin-based chemotherapy regimen was tested in a study of uterine sarcoma in which about 10% of the cases were ESS (¹²⁹). Despite a relatively high response rate of 54%, the regimen was too toxic to be clinically useful. Radiotherapy may help to improve local control, but its role as adjuvant therapy in ESS is unclear (¹³⁰).

Undifferentiated Endometrial Sarcoma

Histology

In undifferentiated endometrial sarcomas, previously considered high-grade ESS, the tumor cells tend to be larger, with more vesicular nuclei, coarse clumps of chromatin, and more prominent nucleoli (¹³¹). Mitotic counts are usually low in low-grade ESS, and this is the principal criterion for distinguishing between the low- and high-grade forms (¹³¹). However, numerous mitotic figures can be present in otherwise-typical low-grade tumors, and high-grade tumors can have only a few mitotic figures (^123,132). Flow cytometry with DNA ploidy seems to be associated more with the behavior of the tumor than with mitotic count (¹³³). In the 2002 classification of ESS by the World Health Organization, high-grade tumors were considered as pleomorphic or undifferentiated sarcoma rather than as a type of ESS because of their distinctive and aggressive clinical behavior. Undifferentiated sarcomas generally stain for Ki67, p16, and p53 and not for estrogen and progesterone receptors. These tumors also do not have any known chromosomal abnormalities (⁵⁸).

Prognosis and Clinical Course

In contrast to low-grade ESS, individuals with undifferentiated sarcomas have poor overall survival regardless of stage. The median progression-free survival, in one series, was 7.3 months, and overall survival was only 11.8 months (¹³⁴). Patients are also likely to progress quickly despite an initial response to treatment (¹³⁴).

Treatment

Surgery

Undifferentiated endometrial sarcomas are treated with up-front surgical management, including hysterectomy and oophorectomy with or without lymphadenectomy. Due to the aggressive nature of this disease, these patients require adjuvant therapy.

Adjuvant Therapy

Unlike ESS, undifferentiated endometrial sarcomas are not responsive to hormone therapy. Adjuvant chemotherapy is similar to that for leiomyosarcoma, as discussed previously.

Adenosarcoma

Histology

Adenosarcomas have two tissue components: The mesenchymal tissue is malignant, and the epithelial part is benign. The epithelial component commonly resembles a proliferative or inactive endometrial gland, often appearing as cleftlike spaces dispersed throughout the proliferative stroma in a phyllode pattern. Slight atypia may be present (⁸¹). The mesenchymal stromal component is usually homologous tissue, such as stromal sarcoma, fibrosarcoma, or leiomyosarcoma (^81,135). The pattern of stromal cell hypercellularity surrounding the glands is characteristic, and the stromal cells show variable atypia and mitosis (⁸⁰). Adenosarcomas with extensive stromal sarcoma proliferation are called adenosarcoma with sarcoma overgrowth if the sarcomatous component constitutes more than 25% of the tumor (^80,81). This subtype should be recognized because its prognosis is worse than that of a typical adenosarcoma and more similar to that of carcinosarcoma (¹³⁶).

Prognosis and Clinical Course

Adenosarcoma is considered a low-grade malignant tumor with an excellent prognosis (^80,81,82). Most cases are confined to the uterus at time of diagnosis and require no adjuvant therapy. Distant metastases have been reported in 5% of cases, and the recurrence rate is approximately 23%, with about one-third of recurrences occurring more than 5 years after initial treatment (^80,81,137). Extrauterine spread at diagnosis, deep infiltrating tumors, sarcomatous overgrowth, and tumor cell necrosis have all been associated with increased risk of recurrence, but again are rare (^{80,104,120,138}).

Treatment

Surgery

Management of adenosarcomas consists of hysterectomy and, as discussed previously, additional therapy is generally not required given the excellent prognosis.

Adjuvant Therapy

Only a few reports have described cases in which adenosarcoma was successfully treated with chemotherapeutic drugs such as liposomal doxorubicin (⁸²). Krivak and colleagues reported adjuvant chemotherapeutic treatment of nine patients with residual or recurrent adenosarcoma with sarcomatous overgrowth. The drugs used were cisplatin and ifosfamide; doxorubicin; and cisplatin and doxorubicin. The progression-free interval for four patients ranged from 7 to 22 months; all patients had died of recurrent or progressive disease by 39 months (¹³⁶).