CHAPTER 33: Endometrial Cancer

In the United States, endometrial cancer is the most common gynecologic malignancy. Risk factors include obesity and advancing age. As these factors are now more prevalent, the incidence of endometrial cancer continues to increase. Fortunately, patients usually seek medical attention early due to vaginal bleeding, and endometrial biopsy leads quickly to diagnosis. The primary treatment is hysterectomy with bilateral salpingo-oophorectomy (BSO) and staging lymphadenectomy for most women. Three quarters will have stage I disease that is curable by surgery alone. Patients with more advanced disease typically require postoperative combination chemotherapy, radiotherapy, or both.

In the United States, women have a 3-percent lifetime risk of developing endometrial cancer. Although an estimated 54,870 new cases were diagnosed, only 10,170 deaths are expected in 2015. As noted, most patients are diagnosed early and subsequently cured. As a result, endometrial cancer is the fourth leading cause of cancer, but the seventh leading cause of cancer deaths among women (Siegel, 2015).

Endometrial adenocarcinomas are categorized as type I or type II based on histology. Type I, that is, endometrioid type, makes up 80 to 90 percent of all cases (Felix, 2010). The other 10 to 20 percent are type II cancers, namely, the non-endometrioid histologic types that include serous and clear cell adenocarcinomas. Risk factors for developing endometrial cancer are numerous (Table 33-1). Risks specifically for type I cancers are associated with an excess-estrogen environment.

Of these, obesity is the most common cause of endogenous overproduction of estrogen. Excessive adipose tissue increases peripheral aromatization of androstenedione to estrone. In premenopausal women, elevated estrone levels trigger abnormal feedback in the hypothalamic-pituitary-ovarian axis. The clinical result is oligo- or anovulation. In the absence of ovulation, the endometrium is exposed to virtually continuous estrogen stimulation without subsequent progestational effect and without menstrual withdrawal bleeding.

Unopposed estrogen therapy is the next most important potential inciting factor. Fortunately, the malignant potential of continuously administered estrogen was recognized more than three decades ago (Smith, 1975). Currently, it is rare to encounter a woman with a uterus who has taken unopposed estrogen for years. Instead, combined estrogen plus progestin hormonal replacement therapy (combination HRT) is routinely prescribed for postmenopausal women with a uterus to reduce estrogen-related endometrial cancer risk (Strom, 2006). Moreover, in one study, the endometrial cancer risk was lower in women taking continuous combination HRT for greater than 6 months compared with those women who had never taken HRT (Phipps, 2011).

Menstrual and reproductive influences are commonly associated with endometrial cancer. For example, early age at menarche or late age of menopause are both associated with increased risk (Wernli, 2006). Classically, women with polycystic ovarian syndrome (PCOS) are anovulatory and thus also have an increased risk of developing this cancer (Fearnley, 2010; Pillay, 2006).

Environment is implicated in endometrial cancer in several ways. Women in Western and developed societies have a much higher incidence (Parkin, 2005). Obvious confounding variables within these populations, such as obesity and low parity, account for much of this effect. However, a possible role for nutrition—especially a diet with a high animal-fat content—is another explanation (Goodman, 1997). Immigrant populations tend to assume the risks of native populations within one or two generations, highlighting the importance of environmental influences (Liao, 2003).

Older age is linked with endometrial cancer development. The average age at diagnosis is the early 60s, and overall, approximately 80 percent of these cancers are diagnosed in postmenopausal women older than 55 years (Madison, 2004; Schottenfeld, 1995). Approximately 8 percent of endometrial cancers develop in patients younger than 45 years (Howlader, 2014). Of note, Nevadunsky and associates (2014) found that the age at diagnosis of endometrioid cancer decreased linearly with increasing body mass index (BMI).

Family history is another risk for endometrial cancer. Endometrial cancer is the most common extracolonic manifestation in Lynch syndrome (hereditary nonpolyposis colorectal cancer [HNPCC]) (Hemminki, 2005). This autosomal-dominant syndrome results primarily from mutations in the mismatch repair genes. The mismatch-repair genes associated with Lynch are MLH1, MSH2, MSH6, and PMS2 (Bansal, 2009). Gene mutation prevents repair of base mismatches, which are commonly produced during DNA replication. Inactivity of this DNA repair system leads to mutations that can promote carcinogenesis. Mutation carriers have a risk of developing endometrial cancer that ranges from 40 to 60 percent. Among affected women, the endometrial cancer risk actually exceeds that for colorectal cancer and often develops at a young age (Aarnio, 1999; Delin, 2004). Of endometrial cancer cases, 2 to 5 percent are attributable to Lynch syndrome (Hampel, 2006). In general, most familial cases develop in premenopausal women (Gruber, 1996).

Women who carry mutations in BRCA1 and BRCA2 genes are at increased risk for breast and ovarian cancer. They may also have a slightly elevated risk for endometrial cancer, but only because associated breast cancers are often treated with tamoxifen (Beiner, 2007; Thai, 1998).

Tamoxifen causes a two- to threefold higher risk of developing endometrial cancer by its modest “unopposed” estrogenic effect on the endometrium (Chap. 27). The increased risk of endometrial cancer affects postmenopausal women almost exclusively, and cancer rates increase linearly with the duration and cumulative dose of tamoxifen therapy (Fisher, 1998; van Leeuwen, 1994). Accordingly, women taking tamoxi­fen are counseled regarding this endometrial risk and should report vaginal spotting, bleeding, or discharge. That said, unless a tamoxifen-treated patient has such symptoms or is identified to be at otherwise high risk for endometrial cancer, routine endometrial surveillance does not increase early detection rates (American College of Obstetricians and Gynecologists, 2014c).

Coexisting medical conditions such as diabetes mellitus, hypertension, and gallbladder disease are more commonly associated with endometrial cancer (Morimoto, 2006; Soliman, 2005). In general, these are frequent sequelae of obesity and an environment of chronic excess estrogen.

In contrast, combination oral contraceptive (COC) use for at least 1 year confers as much as a 30- to 50-percent reduced risk of endometrial cancer, and risk reduction extends for 10 to 20 years (Dossus, 2010; Stanford, 1993). This most likely derives from a chemopreventive effect on the endometrium provided by the progestin component (Maxwell, 2006). Logically, progesterone intrauterine devices (IUDs) also confer long-term endometrial cancer protection (Tao, 2006). Moreover, similar protective effects have been found with inert and copper IUD types (Felix, 2015).

Smokers have a lower risk of developing endometrial cancer. The biologic mechanism is multifactorial but in part involves lower circulating estrogen levels from weight reduction, earlier age at menopause, and altered hormonal metabolism. Both current and past smoking have a long-lasting influence (Viswanathan, 2005).

Most endometrial cancers arise following progression of histologically distinguishable hyperplastic lesions. In fact, endometrial hyperplasia is the only known direct precursor of invasive disease. Endometrial hyperplasia is defined as endometrial thickening with proliferation of irregularly sized and shaped glands and an increased gland-to-stroma ratio (Fig. 33-1) (Ellenson, 2011b). In the absence of such thickening, lesions are best designated as disorderly proliferative endometrium or focal glandular crowding.

Classification

Endometrial hyperplasia represents a continuum of histopathologic findings. The classification system used by the World Health Organization (WHO) and International Society of Gynecological Pathologists designates four different types with varying malignant potential (Table 33-2) (Kurman, 1985, 2014). Hyperplasias are classified as simple or complex, based on the absence or presence of architectural abnormalities of the endometrial glands. Abnormalities include gland crowding and complexity (see Fig. 33-1). Most importantly, hyperplasias are additionally labeled as atypical if they demonstrate nuclear atypia of the endometrial gland cells. Atypical endometrial hyperplasias are clearly associated with the subsequent development of adenocarcinoma. Simple atypical hyperplasia is a relatively uncommon diagnosis. In general, most have a complex architecture.

Although endometrial hyperplasias are formally classified in these four different groups, they tend to be morphologically heterogeneous, both within and between individual patients. This histologic diversity explains why only a small number of conserved features are useful as diagnostic criteria. As a result, reproducible scoring of cytologic atypia is often challenging, particularly with a small amount of tissue from a biopsy sample.

Endometrial intraepithelial neoplasia (EIN) is a term introduced to more accurately distinguish the two very different clinical categories of hyperplasia: (1) normal polyclonal endometria diffusely responding to an abnormal hormonal environment and (2) intrinsically proliferative monoclonal lesions that arise focally and confer an elevated risk of adenocarcinoma (Mutter, 2000). This nomenclature emphasizes the malignant potential of endometrial precancers and is in keeping with similar precedents in the cervix (CIN [cervical intraepithelial neoplasia]), vagina (VaIN), and vulva (VIN) (Chap. 29).

Using this system, anovulatory or prolonged estrogen-exposed endometria without atypia are generally designated as endometrial hyperplasias. In contrast, EIN is used to describe all endometria delineated as premalignant by a combination of three morphometric features. The qualities reflect glandular volume, architectural complexity, and cytologic abnormality. The EIN classification system is a more accurate and reproducible way to predict progression to cancer (Baak, 2005; Hecht, 2005). This classification is endorsed by the Society of Gynecologic Oncology and American College of Obstetricians and Gynecologists (2015) but has not been universally implemented.

Clinical Features and Diagnosis

The risks for developing endometrial hyperplasia generally mirror those for invasive carcinoma (Anastasiadis, 2000; Ricci, 2002). Two thirds of women present with postmenopausal bleeding (Horn, 2004). However, premenopausal women with abnormal uterine bleeding (AUB) are also evaluated as described in Chapter 8.

As hyperplasia is a histologic diagnosis, a Pipelle office endometrial biopsy (EMB) or outpatient dilatation and curettage (D & C) are suitable choices for endometrial sampling. The American College of Obstetricians and Gynecologists (2014a) recommends such sample for women older than 45 years with AUB. EMB is also considered for those younger than 45 with chronic excess estrogen exposure (exogenous or endogenous), failed medical management, and persistent AUB.

In those with AUB, transvaginal sonography to measure endometrial thickness is also a feasible method for predicting endometrial hyperplasia (Granberg, 1991; Jacobs, 2011). In postmenopausal women, endometrial stripe thickness measurements ≤4 mm are associated with bleeding that is attributed to endometrial atrophy (American College of Obstetricians and Gynecologists, 2013). Postmenopausal women with a thicker endometrium warrant biopsy. Sonography may also identify abnormal echostructural changes in the endometrium. Cystic endometrial changes suggest polyps, homogeneously thickened endometrium may indicate hyperplasia, and a heterogeneous structural pattern is suspicious for malignancy (Figs. 33-2 and 33-3). However, these sonographic findings show great overlap and cannot be used alone.

For premenopausal women, transvaginal sonography is often performed to exclude structural sources of abnormal bleeding. Similarly, researchers have attempted to create endometrial thickness guidelines. However, endometrial thicknesses can vary considerably among premenopausal women during normal menstrual cycling. From studies, suggested evidence-based abnormal thresholds range from >4 mm to >16 mm (Breitkopf, 2004; Goldstein, 1997; Shi, 2008). Thus, consensus for an endometrial thickness threshold has not been established for this group. That said, in reproductive-aged women with a thick endometrium combined with other hyperplasia risk factors, EMB may be prudent.

Of other tools, hysteroscopy is more sensitive for focal endometrial lesions. Hyperplastic endometrium is grossly indistinct, and thus hysteroscopy has poor sensitivity for this diagnosis (Ben Yehuda, 1998; Garuti, 2006).

Occasionally, an adnexal mass may be palpable during examination and in most cases is a benign ovarian cyst. However, any solid features noted during transvaginal sonography raises the possibility of a coexisting ovarian granulosa cell tumor. These tumors produce excess estrogen that results in up to a 30-percent risk of endometrial hyperplasia or less commonly, endometrial carcinoma (Chap. 36) (Ayhan, 1994).

Treatment

Management of women with endometrial hyperplasia mainly depends on a patient’s age, comorbid risks for surgery, desire for fertility, and specific histologic features such as cytologic atypia. Traditional treatment has been surgery. Hormonal therapy is another option and includes oral or injectable progestins or the progestin (levonorgestrel-releasing) IUD.

There is some inconsistency of diagnosis and uncertainty in predicting the stability of individual lesions. Specifically, several studies have documented low reproducibility for WHO classifications of endometrial hyperplasia (Allison, 2008; Sherman, 2008; Zaino, 2006). In addition, there is no way to anticipate which types will involute with progestin therapy. However, as long as an endometrial sample is representative and a provider has no reason to suspect a coexisting invasive carcinoma, the decision to treat endometrial hyperplasia through hormonal or surgical means relies on clinical judgment.

Nonatypical Endometrial Hyperplasia

Premenopausal Women

Nonatypical lesions may spontaneous regress without therapy. However, progestins are generally used to address the underlying etiology, that is, chronic anovulation and excess estrogen (Terakawa, 1997). Premenopausal women with nonatypical endometrial hyperplasia typically require a 3- to 6-month course of low-dose progestin therapy. Cyclic medroxyprogesterone acetate (MPA) (Provera) given orally for 12 to 14 days each month at a dose of 10 to 20 mg daily is commonly used. Continuous daily dosing with MPA 10 mg is suitable and may be more effective than cyclic administration in reversing hyperplastic changes. Another frequently used option is COC pills for those without contraindications. The levonorgestrel-releasing IUD is also effective (Gallos, 2010; Ørbo, 2014; Scarselli, 2011).

In general, follow-up endometrial biopsy is performed to document regression. In those with an IUD, endometrial biopsy can be performed without device removal. After regression, a key point is to continue endometrial protection. Thus, once hyperplastic changes resolve, patients are continued on progestins and observed until menopause. Additional endometrial sampling is required for new bleeding.

Postmenopausal Women

Postmenopausal women with nonatypical endometrial hyperplasia may also be treated with low-dose oral cyclic MPA or a continuous 10-mg daily regimen. However, it is particularly important in older women to be confident that the sample obtained is adequate for excluding cytologic atypia. D & C may be indicated in some circumstances, especially if the tissue from Pipelle sampling is scant or if recurrent bleeding is noted.

In most cases, a form of progestin therapy is used to treat endometrial hyperplasia without atypia. But, affected postmenopausal patients who have a contraindication to progestin therapy or who cannot tolerate the therapy can be expectantly managed. Complex hyperplasia without atypia is usually treated chronically with progestins. With either complex or simple hyperplasia without atypia, office endometrial biopsy is recommended every 3 to 6 months until lesion resolution is achieved.

Response to Progestins

In cases of endometrial hyperplasia without atypia, the risk of progression to endometrial cancer is low (1 to 3 percent). The overall clinical and pathologic regression rates to progestin therapy range from 70 to 80 percent for nonatypical endometrial hyperplasia (Rattanachaiyanont, 2005; Reed, 2009). Patients with persistent disease on repeated biopsy may be switched to a higher-dose regimen such as MPA, 40 to 100 mg orally daily. Also, megestrol acetate (Megace), 160 mg daily or 80 mg twice daily, is suitable. It can be increased even up to 160 mg twice daily if no regression is initially achieved. Again, a clinician must confirm that hormonal ablation has occurred by resampling the endometrium after a suitable therapeutic interval, usually 3 to 6 months. Hysterectomy may also be considered for lesions that are refractory to medical management.

If surgery is selected, a minimally invasive surgery (MIS) approach is considered, and options are laparoscopic, robotic, or vaginal hysterectomy. In cases in which hyperplasia has been proven or is suspected, the uterus is removed in toto and without morcellation, which might disseminate the lesion. Because the lesion may extend into the lower uterine segment or upper endocervix, supracervical hysterectomy is not appropriate for women undergoing hysterectomy for treatment of endometrial hyperplasia.

Atypical Endometrial Hyperplasia

Hysterectomy is the preferred treatment for women with atypical endometrial hyperplasia because the risk of progression to cancer over time approximates 29 percent. There is also a high rate of finding concurrent invasive malignancy coexistent with the atypical hyperplasia (Horn, 2004; Trimble, 2006). In postmenopausal women, a hysterectomy with removal of both tubes and ovaries is recommended.

In premenopausal women who have completed childbearing, hysterectomy is performed for atypical hyperplasia. Risk-reducing salpingectomy is encouraged to potentially lower cancer risk that arises from the fallopian tubes (American College of Obstetricians and Gynecologists, 2015d). For premenopausal women, removal of the ovaries is optional. The deciding factors mirror those for women contemplating BSO for other benign indications and are outlined fully in Section 43-12.

Premenopausal women who strongly wish to preserve ferti­lity can be treated with progestins (Trimble, 2012). High-dose progestin therapy, megestrol acetate 80 mg orally twice daily, is an option for motivated patients who will be compliant with surveillance (Randall, 1997). The IUD that releases 20 μg of intrauterine levonorgestrel daily (Mirena) is also suitable (Ørbo, 2014). Poor surgical candidates may also warrant an attempt at hormonal ablation with progestins. Resolution of the hyperplasia must be confirmed by serial endometrial biopsies every 3 months until response is documented. Otherwise, hysterectomy is recommended. Following hyperplasia resolution, surveillance and progestins continue long-term due to the potential for eventual progression to carcinoma (Rubatt, 2005). Once fertility is complete, hysterectomy is again recommended.

The Gynecologic Oncology Group (GOG) performed a prospective cohort study of 289 patients who had a diagnosis of atypical endometrial hyperplasia. Participants underwent hysterectomy within 3 months of their biopsy, and 43 percent were found to have a concurrent endometrial carcinoma (Trimble, 2006). Suh-Burgmann and associates (2009) found a similarly high number of 48 percent. Results demonstrate the difficulty in attaining an accurate diagnosis before hysterectomy and the potential risks of conservative hormonal treatment.

Generalists in obstetrics and gynecology who perform hysterectomy for atypical endometrial hyperplasia should recognize the possibility of a coexisting invasive malignancy and the possible need for surgical staging. At a minimum, peritoneal washings are obtained prior to performing a hysterectomy. In addition, the uterus should be opened and examined in the operating room. From this, a frozen section analysis can be performed to search for concurrent cancer and determine grade and depth of invasion if found. Any suspicion for myometrial invasion is an appropriate indication for intraoperative consultation with a gynecologic oncologist.

Pathogenesis

Endometrial cancer is a biologically and histologically diverse group of neoplasms characterized by a dualistic model of pathogenesis. As noted, type I endometrioid adenocarcinomas comprise most cases. They are estrogen-dependent, low grade, and derived from atypical endometrial hyperplasia. In contrast, type II cancers are serous or clear cell histology, have no precursor lesion, and portend a more aggressive clinical course (Table 33-3). The morphologic and clinical differences are paralleled by genetic distinctions. Namely, type I and II tumors carry mutations of independent sets of genes (Bansal, 2009; Hecht, 2006). The two pathways of endometrial cancer pathogenesis have significant overlap and thus result in a spectrum of histologic features.

Prevention

Education can be effective prevention, as many endometrial cancer risks are alterable. Women with PCOS may benefit from weight loss and chronic progestin supplementation (Chap. 17). Assessing and managing obesity as described in Chapter 1 may also lower risks.

For women at average risk or increased risk, routine screening of hyperplasia or endometrial cancer is not advocated. Instead, at the onset of menopause, women are counseled on the risks and symptoms of endometrial cancer and strongly encouraged to report unexpected bleeding or spotting to their provider. One screening exception is the woman with Lynch syndrome. For these individuals, EMB is recommended every 1 to 2 years beginning at age 30 to 35 years (American College of Obstetricians and Gynecologists, 2014b; Smith, 2015).

Genetic testing criteria have been published to identify the individual with Lynch syndrome (Table 33-4) (Lancaster, 2015). Lynch syndrome cancers include colon, endometrium, small bowel, renal pelvis and ureter, and ovary, among others (Vasen, 1999). Referral for genetic counseling can further clarify which patients may benefit from specific germline testing (Balmana, 2006; Chen, 2006). Endometrial cancer is the most common “sentinel cancer,” thus, obstetrician-gynecologists play a pivotal role in the identification of women with Lynch syndrome (Lu, 2005).

Since women with Lynch syndrome have such a high lifetime risk of developing endometrial cancer (40 to 60 percent), prophylactic hysterectomy is recommended once affected women reach the early to mid 40s. In a cohort of 315 HNPCC-mutation carriers, Schmeler and coworkers (2006) confirmed the benefit of this approach by reporting a 100-percent endometrial cancer risk reduction. In general, BSO is also performed due to the 9- to 12-percent lifetime risk of ovarian cancer. Prior to hysterectomy, colon cancer screening with colonoscopy should be up to date (American College of Obstetrician and Gynecologists, 2014b).

Diagnosis

Signs and Symptoms

Early diagnosis of endometrial cancer is almost entirely dependent on the prompt recognition and evaluation of irregular vaginal bleeding. In premenopausal women, a clinician must maintain a high index of suspicion for a history of prolonged, heavy menstruation or intermenstrual spotting, because many other benign disorders give rise to similar symptoms (Table 8-1). Postmenopausal bleeding is particularly worrisome, leading to a 5- to 10-percent likelihood of diagnosing endometrial carcinoma (Gredmark, 1995; Iatrakis, 1997). Abnormal vaginal discharge may be another symptom in older women.

Unfortunately, some patients do not seek medical attention despite months or years of heavy, irregular bleeding. In more advanced disease, pelvic pressure and pain may reflect uterine enlargement or extrauterine tumor spread. Patients with serous or clear cell tumors often present with signs and symptoms suggestive of advanced epithelial ovarian cancer that include pelvic pain or pressure, bloating, early satiety, and increasing abdominal girth (Chap. 35).

Papanicolaou Test

Pap testing is not an indicated test to diagnose endometrial cancer, and 50 percent of women with endometrial cancer will have normal findings (Gu, 2001). Liquid-based cytology appears to increase the detection of glandular abnormalities, but not enough to cause a shift in clinical practice (Guidos, 2000; Schorge, 2002). However, some findings from Pap testing should prompt further investigation. Benign endometrial cells are occasionally recorded on a routine Pap test report in women 45 years or older. In premenopausal women, this is often a finding of limited importance, especially if a test is obtained following menses. However, postmenopausal women with such findings have a 3- to 5-percent risk of endometrial cancer (Simsir, 2005). In those using HRT, the prevalence of benign endometrial cells on smears is increased, and the associated risk of malignancy is less (1 to 2 percent) (Mount, 2002). Although endometrial biopsy is considered in asymptomatic postmenopausal women if this finding is reported, most patients ultimately diagnosed with hyperplasia or cancers also have concomitant abnormal bleeding (Ashfaq, 2001).

In contrast, atypical glandular cells found during Pap testing carry higher risks for underlying cervical or endometrial neoplasia. Accordingly, evaluation of a glandular abnormality includes colposcopy and endocervical curettage (ECC). It may also include endometrial sampling in nonpregnant patients older than 35 years or in those younger if there is a history of abnormal bleeding, if risk factors for endometrial disease are noted, or if the cytology specifies that the atypical glandular cells are of endometrial origin.

Endometrial Sampling

Office Pipelle biopsy is preferred for the initial evaluation of women with bleeding suspicious for malignancy (Feldman, 1993). However, if sampling techniques fail to provide sufficient diagnostic information or if abnormal bleeding persists, D & C may be required to clarify the diagnosis.

The American College of Obstetricians and Gynecologists (2015b) considers hysteroscopy acceptable for AUB evaluation in those without suspected advanced-stage uterine or cervical cancer. However, hysteroscopy is more sensitive for focal endometrial lesions and thus has proved less helpful in diagnosing early endometrial cancer. In those cases in which hysteroscopy is used to evaluate abnormal bleeding and in which cancer is ultimately diagnosed, an increased incidence of positive peritoneal cytology has been noted during subsequent staging surgery (Obermair, 2000; Polyzos, 2010; Zerbe, 2000). Although the risk of peritoneal contamination by cancer cells may be increased by retrograde efflux of hysteroscopic media, patient prognosis overall does not appear to be worsened (Cicinelli, 2010; Revel, 2004).

Laboratory Testing

The only clinically useful tumor marker in the management of endometrial cancer is a serum CA125 level. Preoperatively, an elevated level indicates the possibility of more advanced disease (Powell, 2005). In practice, it is most useful in patients with advanced disease or serous subtypes to assist in monitoring response to therapy or during posttreatment surveillance. However, even in this setting, its utility in the absence of other clinical findings is limited (Price, 1998).

Imaging Studies

In general, for women with a well-differentiated type I endometrioid tumor, chest radiography is the only required preoperative imaging study. All other preoperative testing is directed toward general surgical preparation (Chap. 39).

Computed tomography (CT) or magnetic resonance (MR) imaging is usually not necessary (American College of Obstetricians and Gynecologists, 2015c). However, CT scanning can be obtained preoperatively in cases with higher-grade lesions to assess for lymph node involvement or metastatic disease. MR imaging can occasionally help distinguish an endometrial cancer with cervical extension from a primary endocervical adenocarcinoma (Nagar, 2006). Moreover, women with serous features or other high-risk histology on preoperative biopsy and those with physical examination findings suggesting advanced disease are most appropriate for abdominopelvic CT scanning (Fig. 33-4). In these cases, advance knowledge of intraabdominal disease may help guide surgery and treatment. MR imaging is also recommended for women who are considering fertility-sparing management with hormonal therapy, since it may not be an option if deep invasion is found.

Role of the Generalist

Although most endometrial cancers are cured by hysterectomy and BSO, primary management by gynecologic oncologists has advantages. It is an efficient use of health care resources, minimizes potential morbidity, is more likely to lead to staging, and improves the survival of patients with high-risk disease (Chan, 2011; Roland, 2004). Therefore, preoperative consultation is generally advisable for any patient with endometrial cancer who is being prepared for surgery by a generalist in obstetrics and gynecology. Postoperatively, a gynecologic oncologist should be consulted if cervical extension, extrauterine disease, or positive peritoneal washing cytology was found during surgery.

If treated by an oncologist, early-stage patients treated by surgery alone will return in many cases to their primary obstetrician-gynecologist for surveillance. Consultation is again re­commended if recurrent disease is later suspected or identified.

When an endometrial cancer is unexpectedly diagnosed after hysterectomy performed by a generalist for other indications, consultation is also recommended. Possible therapeutic options include no further therapy and surveillance only, reoperation to complete surgical staging, or radiotherapy to prevent local recurrence. In general, the survival advantages of staging must be weighed against the complications from another surgical procedure (American College of Obstetricians and Gynecologists, 2015c). Fortunately, the advent of laparoscopic and robotic delayed staging offers the potential for less morbidity (Spirtos, 2005).

Pathology

The spectrum of aggressiveness within the histopathologic types of endometrial cancer is broad (Table 33-5). Most patients have endometrioid adenocarcinomas that behave indolently. However, some will have an unfavorable histology that portends a much more aggressive tumor. In addition, the degree of tumor differentiation is an important predictor of disease spread.

Histologic Grade

The most widely used grading system for endometrial carcinoma is the three-tiered International Federation of Gynecology and Obstetrics (FIGO) system (Table 33-6). Grade 1 lesions typically are indolent with little propensity to spread outside the uterus or recur. Grade 2 tumors have an intermediate prognosis. Grade 3 cancers pose an increased potential for myometrial invasion and nodal metastasis.

Histologic grading is primarily determined by the tumor’s architectural growth pattern (Zaino, 1994). However, there are a few exceptions, and the optimal method for determining grade is somewhat controversial. Nuclear atypia that is inappropriately advanced relative to the architectural grade raises a grade 1 or 2 tumor by one level. For example, a grade 2 lesion based on architectural features may be increased to a grade 3 lesion if significant nuclear atypia is present (Zaino, 1995). Nuclear grading based on the FIGO system is also used for all serous and clear cell adenocarcinomas (Pecorelli, 1999).

Histologic Type

Endometrioid Adenocarcinoma

This is the most common histologic type of endometrial cancer and accounts for more than 75 percent of cases. This type I tumor characteristically contains glands that resemble those of the normal endometrium (Fig. 33-5). The concomitant presence of hyperplastic endometrium correlates with a low-grade tumor and a lack of myometrial invasion. However, when the glandular component decreases and is replaced by solid nests and sheets of cells, the tumor is classified as a higher grade (Kurman, 2014). In addition, an atrophic endometrium is more frequently associated with high-grade lesions that have a greater potential to metastasize (Kurman, 1994).

Endometrioid adenocarcinomas may also display variant forms. These include endometrioid adenocarcinoma with squamous differentiation or with villoglandular or secretory types (Fig. 33-6). In general, the biologic behavior of these variant tumors reflects that of classic endometrioid adenocarcinoma.

Serous Carcinoma

Accounting for 5 to 10 percent of endometrial cancers, serous carcinoma typifies the highly aggressive type II tumors that arise from the atrophic endometrium of older women (Jordan, 2001). There is typically a complex pattern of papillary growth, and cells demonstrate marked nuclear atypia (Fig. 33-7). Commonly referred to as uterine papillary serous carcinoma (UPSC), its histologic appearance resembles epithelial ovarian cancer, and psammoma bodies are seen in 30 percent of cases (Kurman, 2014).

Grossly, the tumor is exophytic with a papillary appearance emerging from a small, atrophic uterus. These tumors may occasionally be confined within a polyp and have no evidence for spread (Carcangiu, 1992). However, UPSC has a known propensity for myometrial and lymphatic invasion. Intraperitoneal spread, such as omental caking, which is unusual for typical endometrioid adenocarcinoma, is also common even when myometrial invasion is minimal or absent (Fig. 33-8) (Sherman, 1992). As a result, it may be impossible to distinguish UPSC from epithelial ovarian cancer during surgery. Like ovarian carcinoma, these tumors usually secrete CA125. Thus, serial serum measurements can be useful marker to monitor disease postoperatively. UPSC is an aggressive cell type, and women with mixed endometrial cancers containing as little as 25 percent of UPSC have the same survival as those with pure uterine serous carcinoma (Ellenson, 2011a).

Clear Cell Carcinoma

Fewer than 5 percent of endometrial cancers are clear cell variants, but this is the other major type II tumor (Abeler, 1991). The microscopic appearance may be predominantly solid, cystic, tubular, or papillary. Most frequently, it consists of a mixture of two or more of these patterns (Fig. 33-9).

Endometrial clear cell adenocarcinomas are similar to those arising in the ovary, vagina, and cervix. Grossly, there are no characteristic features, but like UPSC, they tend to be high-grade, deeply invasive tumors. Patients are often diagnosed with advanced disease and have a poor prognosis (Hamilton, 2006).

Mucinous Carcinoma

One to 2 percent of endometrial cancers have a mucinous appearance that forms more than half of the tumor. However, many endometrioid adenocarcinomas will have this as a focal component (Ross, 1983). Typically, mucinous tumors have a glandular pattern with uniform columnar cells and minimal stratification (Fig. 33-10). Almost all are stage I grade 1 lesions and carry a good prognosis (Melhem, 1987). Since endocervical epithelium merges with the lower uterine segment, the main diagnostic dilemma is differentiating this tumor from a primary cervical adenocarcinoma. Immunostaining may be helpful, and MR imaging may further clarify the most likely site of origin. In general, to define anatomy, MR imaging is preferable to CT scanning as MR offers superior contrast resolution at soft-tissue interfaces.

Mixed Carcinoma, Undifferentiated Carcinoma, and Rare Types

An endometrial cancer may demonstrate combinations of two or more pure types. To be classified as a mixed carcinoma, a component must make up at least 10 percent of the tumor. Except for serous and clear cell histology, the combination of other types usually has no clinical significance. As a result, mixed carcinoma usually refers to an admixture of a type I (endometrioid adenocarcinoma and its variants) and a type II carcinoma (Kurman, 2014).

Undifferentiated carcinoma lacks architectural differentiation and is characterized by medium-sized, monotonous epithelial cells growing in solid sheets without a pattern (Silva, 2007). These represent 1 to 2 percent of endometrial cancers. Overall, the prognosis is worse than in patients with poorly differentiated endometrioid adenocarcinomas (Altrabulsi, 2005).

Of rare histologic types, fewer than 100 cases of squamous cell carcinoma of the endometrium have been reported. Diagnosis requires exclusion of an adenocarcinoma component and no connection with the squamous epithelium of the cervix (Varras, 2002). Typically, the prognosis is poor (Goodman, 1996). Transitional cell carcinoma of the endometrium is also rare, and metastatic disease from the bladder or ovary must be excluded during diagnosis (Ahluwalia, 2006).

Patterns of Spread

Endometrial cancers have several different potential ways to spread beyond the uterus (Morrow, 1991). Type I endometrioid tumors and their variants most commonly spread, in order of frequency, by: (1) direct extension, (2) lymphatic metastasis, (3) hematogenous dissemination, and (4) intraperitoneal exfoliation. Type II serous and clear cell carcinomas have a particular propensity for extrauterine disease, in a pattern that closely resembles epithelial ovarian cancer. In general, the various patterns of spread are interrelated and often develop simultaneously.

Invasion of the endometrial stroma and exophytic expansion within the uterine cavity follows initial growth of an early cancer. Over time, the tumor invades the myometrium and may ultimately perforate the serosa (Table 33-7). Tumors situated in the lower uterine segment tend to involve the cervix early, whereas those in the upper corpus tend to extend to the fallopian tubes or serosa. Advanced regional growth may lead to direct invasion into adjacent pelvic structures, including the bladder, large bowel, vagina, and broad ligament.

Lymphatic channel invasion and metastasis to the pelvic and paraaortic nodal chains can follow tumor penetration of the myometrium (Table 33-8). The lymphatic network draining the uterus is complex, and patients can have metastases to any single nodal group or combination of groups (Burke, 1996). This haphazard pattern is in contrast to cervical cancer, in which lymphatic spread usually follows a stepwise progression from pelvic to paraaortic to scalene nodal groups.

Hematogenous dissemination most commonly results in metastases to the lung and less commonly to the liver, brain, bone, and other sites. Deep myometrial invasion is the strongest predictor of this pattern of spread (Mariani, 2001a).

Retrograde transtubal transport of exfoliated endometrial cancer cells carries malignant cells to the peritoneal cavity. Serosal perforation of the tumor is another possible pathway. Most types of endometrial cancer cells found in the peritoneal cavity disappear within a short time and have low malignant potential (Hirai, 2001). Alternatively, in the presence of other high-risk features, such as adnexal metastases or serous histology, widespread intraabdominal disease may result.

Port-site metastasis is a rare but possible method of cancer spread. Martínez and coworkers (2010) evaluated nearly 300 laparoscopic staging procedures for endometrial cancer. Port-site metastases complicated 0.33 percent of cases. Similarly, cancer dissemination following specimen morcellation has been reported (Graebe, 2015).

Treatment

Surgical Management

Patients with endometrial cancer should undergo hysterectomy, BSO, and surgical staging (including pelvic washings and lymphadenectomy) using the revised FIGO system (Table 33-9 and Fig. 33-11) (Mutch, 2009). For optimal patient management, the histopathologic description of the preoperative biopsy findings is carefully reviewed. Almost three quarters of patients are stage I at diagnosis (Table 33-10). Only a few circumstances contraindicate primary surgery and include a desire to preserve fertility, massive obesity, high operative risk, and clinically unresectable disease. In general, an extrafascial hysterectomy, also known as type I or simple hysterectomy, is sufficient. However, radical hysterectomy (type III hysterectomy) may be preferable for patients with clinically obvious cervical extension of endometrial cancer (Cornelison, 1999; Mariani, 2001b). Differences in these hysterectomy types are outlined in Table 30-7. Vaginal hysterectomy with or without BSO is another option for women who cannot undergo systematic surgical staging due to comorbidities (American College of Obstetricians and Gynecologists, 2015c). Previously, laparotomy had been the standard approach. However, laparoscopic and robotic surgical staging are increasingly used for endometrial cancer that appears clinically confined to the uterus. Such MIS staging is safe, feasible, and now recommended (Walker, 2009).

Regardless of approach, upon entering the peritoneal cavity, washings are obtained by pouring 50 to 100 mL of sterile saline into the pelvis and collecting it for cytologic assessment. Retrieval of ascitic fluid is an acceptable alternative, but ascites is infrequently encountered. Next, a thorough intraabdominal and pelvic exploration is performed, and suspicious lesions are biopsied or excised. These preliminary procedures are followed by hysterectomy and BSO. The uterus is opened away from the operating table, and the depth of myometrial penetration may be determined by intraoperative gross examination or microscopic frozen section analysis (Sanjuan, 2006; Vorgias, 2002).

As noted, the risk of lymph node metastasis correlates with the tumor grade and depth of invasion into the myometrium. Historically, the combination of preoperative biopsy grade and intraoperative assessment of the depth of myometrial invasion were the two factors that surgeons used to determine whether to proceed with pelvic and paraaortic lymph node dissection. The inaccuracy of this approach has been reported (Eltabbakh, 2005; Leitao, 2008; Papadia, 2009). In addition, the depth of myometrial invasion determined in the operative room is often inaccurate (Frumovitz, 2004a,b).

Following hysterectomy and BSO, concurrent lymphadenectomy allows detection of positive nodes to guide appropriate treatment. Some retrospective studies have shown improved survival rates in patients who had undergone an adequate lymph node dissection (Kilgore, 1995; Todo, 2010). However, the advantage appears to be confined to those in high-risk groups. Thus, although still with some controversy, complete surgical staging with pelvic and paraaortic lymphadenectomy is recommended for patients with high-risk grade 1 endometrioid cancer and any case of grade 2, grade 3 or type II cancers. Lymph nodal staging for cases of low-risk grade 1 endometrioid cancer also is debated (Miller, 2006). Authors in two randomized trials reported no improvement in disease-free or overall survival rates after lymphadenectomy in patients with early-stage disease (Benedetti Panici, 2008; Kitchener, 2009). However, these trials were criticized because lymph node counts were low and lymph node status did not dictate treatment, as many patients received postoperative radiation regardless of their lymph node status. Moreover, concern exists that omitting lymphadenectomy may lead to missed metastatic disease and subsequent insufficient postoperative treatment. In addition, microscopic nodal disease may be unknowingly resected during lymphadenectomy and prevent future relapse.

Distinct from the above discussion, any suspicious pelvic or paraaortic lymph nodes should be removed and histologically evaluated. Excision of grossly involved lymph nodes leads to a survival advantage (Havrilesky, 2005).

Those patients with serous or clear cell features on preoperative biopsy should have extended surgical staging with an infracolic omentectomy and bilateral peritoneal biopsies of the pelvis, pericolic gutter, and diaphragm (Bristow, 2001a). As in ovarian cancer, a surgeon is also prepared to resect any metastases (Bristow, 2000).

Sentinel lymph node evaluation, as done in vulvar and breast cancers, is being investigated and may become a useful technique in endometrial cancer (Abu-Rustum, 2009). The practice does vary but usually involves injecting the ectocervix prior to surgery at the 3, 6, 9 and 12 o’clock positions with technetium sulfur colloid. Subsequent lymphoscintigraphy guides one to the sentinel lymph nodes. At the time of surgery, the cervix is also injected with lymphazurin or isosulfan blue at the 3 and 9 o’clock position. The radioactive and blue nodes are then identified (Frati, 2015). Although investigational, sentinel lymph node evaluation is appealing in that it reduces the potential side effects associated with complete lymphadenectomy, including lymphedema and lymphocyst formation, longer operating times, and increased blood loss. The sensitivity of sentinel lymph node mapping appears acceptable in small series, and a prospective study is being considered by the GOG.

As noted, an MIS approach is acceptable for suitable candidates undergoing hysterectomy and staging for endometrial cancer (Childers, 1994; Spirtos, 2005). The GOG LAP2 study was the first multicenter randomized trial of laparoscopy to address this. In the trial, conventional surgery including pelvic/paraaortic lymphadenectomy was compared with the same steps completed laparoscopically for clinical stage I and IIA endometrial carcinoma. Investigators found laparoscopic staging to be feasible and safe (Walker, 2009). Laparoscopy was completed without conversion in 74 percent of patients randomized to MIS. Advantageously, compared with those undergoing laparotomy, patients undergoing laparoscopy had similar rates of intraoperative injuries (9 versus 8 percent), fewer moderate to severe complications (14 versus 21 percent), shorter hospital stays (median 3 versus 4 days), and better quality of life at 6 weeks postoperatively. However, laparoscopic staging was linked with longer operative times (Kornblith, 2009; Walker, 2009). Long-term treatment success is not compromised with laparoscopic staging, and overall survival and recurrence rates in early reports are similar to those for a traditional abdominal approach (Ghezzi, 2010; Magrina, 1999; Walker, 2012; Zullo, 2009).

Robot-assisted laparoscopic staging of endometrial cancer has been embraced by many gynecologic oncologists to overcome MIS technical challenges, especially in obese patients. Early evidence shows it to be feasible and safe (Hoekstra, 2009). Compared with a laparoscopic approach for endometrial cancer staging, both major complication rates and mean number of lymph nodes removed are similar. The robotic approach results in less blood loss (Cardenas-Goicoechea, 2010; Seamon, 2009).

As described in Chapter 41, not all women are candidates for MIS. Limiting factors can include extensive adhesive disease, a large bulky uterus, morbid obesity, cardiopulmonary disease, and other patient comorbidities. Importantly, morcellation is avoided in cancer cases to prevent disease spread.

Surveillance

Most surgically treated patients can simply be followed by pelvic examination every 3 to 6 months for the first 2 years and then every 6 to 12 months thereafter (National Comprehensive Cancer Network, 2015). Pap testing is not a mandatory part of surveillance, as an asymptomatic vaginal recurrence is identified in less than 1 percent of patients (Bristow, 2006a; Cooper, 2006).

Women who have more advanced cancer that requires postoperative radiation or chemotherapy or both warrant more aggressive monitoring. Serum CA125 measurements may be valuable, particularly for UPSC, if the level was elevated prior to treatment. Intermittent imaging using CT scanning or MR imaging may also be indicated. In general, the pattern of recurrent disease depends on the original sites of metastasis and the treatment received.

Chemotherapy

Only three cytotoxic drugs with definite activity for endometrial cancer have been identified. Paclitaxel (Taxol), doxorubicin (Adriamycin), and cisplatin (Platinol) form TAP chemotherapy, which is one of the adjuvant treatment options for advanced endometrial cancer following surgery. In one GOG trial of 273 women (protocol #177), administration of seven courses of TAP was superior to doxorubicin plus cisplatin, but toxicity was increased—particularly peripheral neuropathy (Fleming, 2004). A less toxic alternative to TAP chemotherapy is paclitaxel plus carboplatin. Routinely used for ovarian cancer, this regimen is effective in advanced-stage endometrial cancer (Hoskins, 2001; Sovak, 2006, 2007). One GOG trial (protocol #209) compared TAP and the regimen of carboplatin plus paclitaxel. Results demonstrated that carboplatin plus paclitaxel was not inferior to TAP in terms of progression-free and overall survival rates. The toxicity profile favored carboplatin plus paclitaxel, and this regimen is generally used at our institution (Miller, 2012).

In practice, cytotoxic chemotherapy is frequently combined, sequenced, or sandwiched with radiotherapy in patients with advanced endometrial cancer following surgery. To reduce toxicity, directed pelvic or paraaortic radiation is usually employed rather than whole abdominal irradiation (Homesley, 2009; Miller, 2009).

Targeted Therapy

In some cases, current therapies fail to provide long-term disease control, and thus novel treatments are being investigated for endometrial cancer. Bevacizumab (Avastin), a vascular endothelial growth factor (VEGF) inhibitor, blocks angiogenesis. In a phase II trial for recurrent/persistent endometrial cancer, this targeted agent showed some salutary action, but further study is needed (Aghajanian, 2011). Another pathway, the fibroblast growth factor receptor pathway, is a potential target (Lee, 2014). The mTOR pathway is an additional target, as mutations in this pathway may lead to endometrial cancer. The GOG (protocol #0286B) is currently studying metformin with chemotherapy for the treatment of advanced/recurrent endometrial cancer. Early data show that metformin inhibits cell proliferation in endometrial cancer cell lines and that this effect is partially mediated through mTOR pathway inhibition. In addition, treatment with metformin in combination with paclitaxel resulted in a synergistic antiproliferative effect in these cell lines.

Radiation

Primary Radiation Therapy

This option rather than surgery is selected rarely and mainly for exceptionally poor surgical candidates. Intracavitary brachytherapy such as Heyman capsules with or without external beam pelvic radiation is the typical method (Chap. 28). In general, the survival rate is 10 to 15 percent lower than that with surgical treatment (Chao, 1996; Fishman, 1996). These poor results suggest that a careful preoperative evaluation and appropriate consultation should be completed before any woman is denied the benefits of hysterectomy (American College of Obstetricians and Gynecologists, 2015c).

Adjuvant Radiation Therapy

In general, this option is offered after staging surgery to women at risk for endometrial cancer recurrence. Those with low-risk early-stage cancer are typically adequately treated with surgery, may not benefit from adjuvant therapy, and usually simply begin surveillance. The use of radiation in early-stage disease has been evaluated in three major trials, all of which demonstrated that adjuvant radiation improved local disease control and recurrence-free survival rates but did not decrease the rate of distant metastases or improve overall survival rates at 5 years (Aalders, 1980; Creutzberg, 2001, 2004; Keys, 2004).

However, in one of these trials, the recurrence-rate reduction was particularly evident in a high-intermediate risk subgroup of women: (1) with three risk factors (grade 2 or 3 tumors, lymphovascular invasion, and invasion of the outer third of the myometrium); (2) with age >50 years and two of these risk factors; and (3) with age >70 years and one risk factor (Keys, 2004). These elements have found their way into both clinical management and the design of more contemporary trials. Therefore, adjuvant radiation is usually offered to patients with high-intermediate risk stage I uterine cancers. However, because of the results of subsequent trials, the type of radiation offered is vaginal brachytherapy. The PORTEC-2 trial showed that vaginal brachytherapy was equivalent to whole pelvic radiation in those patients with high-intermediate risk endometrial cancer (Nout, 2010). Of note, local radiation can reduce the risk of local recurrence but does not improve overall survival rates.

For women with stage II endometrial cancer, the efficacy of postoperative radiotherapy is even harder to decipher. Most data derive from retrospective, single-institution experiences, and evidence supports external beam pelvic radiation, vaginal brachytherapy, both, or no further treatment (Cannon, 2009; Elshaikh, 2015; Rittenberg, 2005). As such, there is no standard approach, and most patients are treated individually based on coexisting risk factors (Feltmate, 1999).

For most women with stage III endometrial cancer, chemotherapy and/or tumor-directed postoperative external beam radiation is indicated (Barrena-Medel, 2009; Homesley, 2009). Most commonly, radiation therapy is specifically directed at pelvic disease but may extend to the paraaortic area if meta­stases are detected. Currently, results are pending from a randomized trial that compares chemoradiation and chemotherapy (carboplatin and paclitaxel) for advanced endometrial cancer (protocol #258).

Few patients with stage IV disease are candidates for radiotherapy with curative intent. Infrequently, a locally confined stage IVA tumor may be an exception. With stage IV disease, intraperitoneal metastases most often lie outside a tolerated radiation field. Therefore, whole abdominal irradiation is not generally preferable to chemotherapy (Randall, 2006). As a result, the role of radiotherapy is generally palliative in these women (Goff, 1994). Chemotherapy is a reasonable option for patients with advanced endometrial cancer.

Hormonal Therapy

Cancer Treatment

One of the unique characteristics of endometrial cancer is its hormonal responsiveness. Thus, for women who are not surgical candidates, continuous chronic progestin treatment or a levonorgestrel-releasing IUD can be primary treatment (Dhar, 2005; Montz, 2002). Also, in young premenopausal patients who desire fertility, similar primary progestin therapy can be used to reverse pathology.

As adjuvant therapy, single-agent high-dose progestins also have activity in women with advanced or recurrent disease (Lentz, 1996; Thigpen, 1999). Tamoxifen upregulates progesterone-receptor expression and is postulated to thereby improve progestin therapy efficacy. Clinically, high response rates have been noted with tamoxifen used adjunctively with progestin therapy (Fiorica, 2004; Whitney, 2004). In general, toxicity is low, but this combination is most often used for recurrent disease.

Estrogen Replacement Therapy

Due to the presumed role of excess estrogen in endometrial cancer development, estrogen supplementation following endometrial cancer treatment is often met with concern for stimulating malignancy recurrence. However, this effect has not been observed (Suriano, 2001). The GOG attempted to determine the risk of estrogen replacement therapy in endometrial cancer survivors by randomly assigning 1236 women who had undergone surgery for stage I and II endometrial cancer to receive either estrogen or placebo. Although the study did not meet its enrollment goals, the low recurrence rate (2 percent) was promising (Barakat, 2006). Women should be individually counseled regarding risks and benefits before beginning posttreatment estrogen replacement for menopausal symptoms.

Uterine Papillary Serous Carcinoma Management

This most aggressive type of endometrial carcinoma is uncommon, and thus, randomized trials are difficult to perform. As a result, most data are single-institution, retrospective analyses. Treatment is usually individualized but is often different from typical endometrioid adenocarcinoma.

If a preoperative biopsy demonstrates serous features, comprehensive surgical staging for UPSC is recommended. This includes total hysterectomy, BSO, peritoneal washings, pelvic/paraaortic lymph node dissection, infracolic omentectomy, and peritoneal biopsies (Chan, 2003). Even noninvasive disease is often widely metastatic (Gehrig, 2001). Fortunately, patients tend to have a good prognosis if surgical staging confirms that disease is confined to the uterus (stage I/II) (Grice, 1998).

Occasionally, no residual UPSC is evident on the hysterectomy specimen, or the tumor minimally involves the tip of a polyp. These women with surgical stage IA can safely be observed. However, all other patients with stage I disease are considered for adjuvant treatment. For this, one effective strategy is postoperative paclitaxel and carboplatin chemotherapy for three to six cycles combined with concomitant vaginal brachytherapy (Dietrich, 2005; Kelly, 2005). However, some data suggest an intrinsic radioresistance for UPSC tumors (Martin, 2005). In addition, based on the largest reported retrospective review of surgical stage I patients, Huh and coworkers (2003) questioned the benefit of any radiation therapy.

Women with stage II UPSC are more likely to benefit from pelvic radiotherapy with or without chemotherapy following surgery. Those having stage III disease are especially prone to have recurrent disease at distant sites. Accordingly, paclitaxel and carboplatin is considered in addition to tumor-directed radiotherapy after surgery (Bristow, 2001a; Slomovitz, 2003).

In practice, many patients will have stage IVB disease. Aggressive surgical cytoreduction is perhaps most important, because one of the strongest predictors of overall survival is the amount of residual disease. Postoperatively, at least six cycles of paclitaxel and carboplatin chemotherapy are indicated (Barrena-Medel, 2009; Bristow, 2001b; Moller, 2004). Enrollment in a clinical trial is strongly considered for cases of advanced uterine cancer.

Fertility-sparing Management

Hormonal therapy without hysterectomy is an option in carefully selected young women with endometrial cancer who desperately wish to preserve their fertility. Currently, patients considering fertility-sparing treatment are recommended to undergo a diagnostic hysteroscopy, sampling by D & C, and imaging to exclude deep myometrial invasion or extrauterine disease (Burke, 2014). Careful selection is also aided by a reproductive endocrinology consultation that clarifies the patient’s posttreatment conception chances. Importantly, many of the biologic processes that lead to endometrial cancer also contribute to decreased fertility. In general, this strategy should apply only to those with grade 1 (type I tumor) adenocarcinomas and with no imaging evidence of myometrial invasion. Rarely, women with grade 2 lesions may be considered candidates, although it may advisable to further assess their disease laparoscopically (Morice, 2005). The aim of hormonal treatment is to reverse the lesion. However, any type of medical management obviously involves inherent risk that a patient must be willing to accept (Yang, 2005).

Progestins are most commonly used for conservative treatment. Oral megestrol acetate, 160 mg given daily or 80 mg twice daily, can promote cancer regression. Alternatively, oral or intramuscular MPA may be delivered at varying doses (Gotlieb, 2003). The levonorgestrel-releasing IUD is another acceptable option. Combining progestin therapy with tamoxifen or with gonadotropin-releasing hormone agonists is less frequently done (Wang, 2002). Regardless of the hormonal agent, recurrence rates are high during long-term observation (Gotlieb, 2003; Niwa, 2005). Despite the feasibility of this option, dosages, therapy duration, and surveillance schedules are not specifically defined.

Women receiving fertility-sparing management are carefully monitored by repeated endometrial biopsy or D & C every 3 months to assess treatment efficacy. Evidence for persistence often prompts a change in agent or a dosage increase. Hysterectomy and operative staging is recommended if a lesion fails to regress with hormonal therapy or if disease progression is suspected.

Delivery of a healthy infant is a reasonable expectation for those patients who respond to treatment and have normal histologic findings in surveillance endometrial samplings. However, assisted reproductive technologies may be required to achieve pregnancy in some cases. Postpartum, patients are again regularly monitored for recurrent endometrial adenocarcinoma (Ferrandina, 2005). In general, women should undergo hysterectomy at completion of childbearing or whenever the preservation of fertility is no longer desired.

Prognostic Factors

Many clinical and pathologic factors influence the likelihood of endometrial cancer recurrence and survival (Table 33-11) (Lurain, 1991; Schink, 1991). Of these, FIGO surgical stage is the most important overriding variable because it incorporates many of the most important risk factors (Table 33-12). Metastatic disease to the adnexa, pelvic/paraaortic lymph nodes, and peritoneal surfaces is reflected by the FIGO stage.

Recurrent Disease

Patients with recurrent endometrial cancer typically require individualized treatment. In general, the site of relapse is the most important predictor of survival. Depending on the circumstances, surgery, radiation, chemotherapy, or a combination of these may be the best strategy. The most curable scenario is an isolated relapse at the vaginal apex in a previously unradiated patient. These women are usually effectively treated by external beam pelvic radiotherapy. In patients who were previously irradiated, exenteration is often the only curative option (Section 46-4) (Barakat, 1999; Morris, 1996). Nodal recurrences or isolated pelvic disease is more likely to result in further disease progression, regardless of treatment modality. However, either is often an appropriate indication for external beam radiotherapy in those not previously irradiated. Salvage cytoreductive surgery may also be beneficial in selected patients (Awtrey, 2006; Bristow, 2006b).

Widely disseminated endometrial cancer or a relapse not amenable to radiation or surgery is an indication for systemic chemotherapy (Barrena-Medel, 2009). Patients are ideally enrolled in an experimental trial due to the limited duration of response with current salvage regimens and the urgent need for more effective therapy. Paclitaxel and carboplatin is an active combination for recurrent disease and found not to be inferior to TAP (Miller, 2012). Progestin therapy with or without tamoxifen is a less toxic option that is particularly useful in selected cases (Fiorica, 2004; Whitney, 2004). In general, effective palliation of women with incurable, recurrent endometrial cancer requires an ongoing dialogue to achieve the optimal balance between symptomatic relief and treatment toxicity.