Uterine Cancer
BS. Nguyễn Hồng Anh
KEY POINTS
1 The most common risk factors for the development of endometrial carcinoma are related to prolonged, unopposed estrogen stimulation.
2 Office endometrial aspiration biopsy is the accepted first step in evaluating a woman with abnormal uterine bleeding or suspected endometrial pathology.
3 Serous and clear cell endometrial carcinomas make up less than 10% of endometrial cancers, yet account for more than one-half of all endometrial cancer deaths.
4 Most patients with endometrial cancer should undergo surgical staging, including hysterectomy, bilateral salpingo-oophorectomy, and peritoneal cytology. Lymph node assessment is necessary for most patients, but may be omitted in patients with negligible risk of lymphatic spread.
5 The most important adverse prognostic variables in endometrial cancer are advancing patient age, nonendometrioid or grade 3 histology, deep myometrial invasion, lymph-vascular invasion, large tumor size, cervical extension, lymph node metastasis, and intraperitoneal spread.
6 Postoperative adjuvant radiotherapy and chemotherapy in selected patients with endometrial cancer decreases the risk of local vaginal/pelvic recurrence and improves disease-free survival.
7 Overall 5-year survival rate in endometrial cancer is approximately 75%.
8 Radiotherapy is the best treatment option for patients with isolated local–regional recurrences who have not received prior radiation. Isolated vaginal recurrences may be salvaged in up to 80% of patients.
9 Uterine sarcomas are, in general, the most malignant group of uterine tumors and differ from endometrial cancers with regard to risk factors, diagnosis, clinical behavior, pattern of spread, and management.
Endometrial carcinoma is the most common malignancy of the female genital tract, accounting for almost one-half of all gynecologic cancers in the United States. In 2018, an estimated 63,230 new cases and 11,350 cancer-related deaths are anticipated (1). Endometrial carcinoma is the fourth most common cancer, ranking behind breast, lung, and colorectal cancers, and the sixth leading cause of death from malignancy in women. Overall, about 2.8% of women develop endometrial cancer during their lifetimes. Although endometrial carcinoma usually presents as early-stage disease and often is managed without radical surgery or radiotherapy, deaths from endometrial carcinoma rose by about 2% per year from 2010 to 2014. Endometrial cancer is a disease that occurs primarily in postmenopausal women and is increasingly virulent with advancing age.
The definite role of estrogen in the development of most endometrial cancers is established. Any factor that increases exposure to unopposed estrogen increases the risk for endometrial cancer.
EPIDEMIOLOGY AND RISK FACTORS
There appear to be two pathogenetic types of endometrial cancer (2). Type I, accounting for about 75% to 85% of cases, occurs in younger, perimenopausal women with a history of exposure to unopposed estrogen, either endogenous or exogenous. In these women, tumors begin as hyperplastic endometrium and progress to carcinoma. These “estrogen-dependent” tumors tend to be better differentiated and have a more favorable prognosis than tumors not associated with hyperestrogenism. Type II endometrial carcinoma occurs in women without estrogenic stimulation of the endometrium. These spontaneously occurring cancers are not pathologically associated with endometrial hyperplasia, but may arise in a background of atrophic endometrium.
They are less differentiated and associated with a poorer prognosis than estrogendependent tumors. These “estrogen-independent” tumors tend to occur in older, postmenopausal, thin women and are present disproportionately in African American and Asian women. Over the past decade, molecular genetic studies showed that these two tumor types evolve via distinct pathogenetic pathways (3) (see Types I and II Endometrial Carcinoma: Molecular Aberrations, below).
Several risk factors for the development of endometrial cancer have been identified, [1] most of which are related to prolonged, unopposed estrogen stimulation of the endometrium (Table 37-1) (3–5). Nulliparous women have two to three times the risk of parous women. Infertility and a history of irregular menses as a result of anovulatory cycles (prolonged exposure to estrogen without sufficient progesterone) increase risk. Natural menopause occurring after age 52 years increases the risk for endometrial cancer 2.4-fold compared with women who experienced menopause before 49 years of age, probably as a result of prolonged exposure of the uterus to progesterone-deficient menstrual cycles. The risk of endometrial cancer is increased 1.5 times for overweight and more than 2.5-fold for obese women (resulting from excess estrone as a result of conversion of adrenally derived androstenedione by aromatization in peripheral fat) (6). The obesity epidemic in Western countries, together with growing rates of insulin resistance and metabolic syndrome, is expected to increase the incidence of endometrial cancer in coming years.
Table 37-1 Risk Factors for Endometrial Cancer
Characteristic Relative Risk
Nulliparity 2–3
Late menopause 2.4
Obesity
236221–50 lb overweight 3
>50 lb overweight 10
Diabetes mellitus 2.8
Unopposed estrogen therapy 4–8
Tamoxifen therapy 2–3
Atypical endometrial hyperplasia 8–29
Lynch II syndrome 20
Other factors leading to long-term estrogen exposure, such as polycystic ovary syndrome and functioning ovarian tumors, are associated with an increased risk for endometrial cancer. Menopausal estrogen therapy without progestins increases the risk of endometrial cancer four- to eightfold. This risk increases with higher doses and with more prolonged use and can be reduced to essentially baseline levels by the addition of at least 10 days of progestin treatment (7). The use of the antiestrogen tamoxifen for treatment of breast cancer is associated with a two- to threefold increased risk for the development of endometrial cancer, although this finding is confounded by the apparent greater risk of endometrial cancer in women who have breast cancer, with or without treatment with tamoxifen (5). Analysis of diabetes mellitus is confounded by obesity, but is thought to be associated with a modest increase in endometrial cancer risk even after adjusting for other factors (8). Women with Lynch II syndrome (previously referred to as hereditary nonpolyposis colorectal cancer syndrome, or HNPCC), a cancer susceptibility syndrome with germline mutations in mismatch repair genes MLH1, MSH2, and MSH6, have a 40% to 60% lifetime risk for endometrial and colon cancer (9). Other medical conditions, such as hypertension and hypothyroidism, are associated with endometrial cancer, but a causal relationship has not been confirmed.
ENDOMETRIAL HYPERPLASIA
Endometrial hyperplasia represents a spectrum of morphologic and biologic alterations of the endometrial glands and stroma, ranging from an exaggerated physiologic state to carcinoma in situ. Clinically significant hyperplasias usually evolve within a background of proliferative endometrium as a result of protracted estrogen stimulation in the absence of progestin influence.
Endometrial hyperplasias are important clinically because they may cause abnormal bleeding, be associated with estrogen-producing ovarian tumors, result from hormonal therapy, and precede or occur simultaneously with endometrial cancer.
The classification scheme endorsed by the International Society of Gynecological Pathologists is based on architectural and cytologic features and long-term studies that reflect the natural history of the lesions (10) (Table 37-2).
Architecturally, hyperplasias are either simple or complex, differentiated by complexity and crowding of the glandular elements. Simple hyperplasia is characterized by dilated or cystic glands with round to slightly irregular shapes, an increased glandular-to-stromal ratio without glandular crowding, and no cytologic atypia. Complex hyperplasia has architecturally complex (budding and infolding) crowded glands, with less intervening stroma without atypia. Atypical hyperplasia refers to cytologic atypia and can be categorized as simple or complex, depending on the corresponding glandular architecture. Criteria for cytologic atypia include large nuclei of variable size and shape that have lost polarity, increased nuclear-to-cytoplasmic ratios, prominent nucleoli, and irregularly clumped chromatin with parachromatin clearing (Fig. 37-1).
Table 37-2 Classification of Endometrial Hyperplasias
Type of Hyperplasia Progression to Cancer (%)
Simple (cystic without atypia) 1
Complex (adenomatous without atypia) 3
Atypical
Simple (cystic with atypia) 8
Complex (adenomatous with atypia) 29
From Kurman RJ, Kaminski PF, Norris HJ. The behavior of endometrial hyperplasia: a long term study of “untreated” hyperplasia in 170 patients. Cancer 1985;56:403–412, with permission. The risk of endometrial hyperplasia progressing to carcinoma is related to the presence and severity of cytologic atypia. Kurman et al. retrospectively studied endometrial curettings from 170 patients with untreated endometrial hyperplasia followed for a mean of 13.4 years (11). They found that progression to carcinoma occurred in 1% of patients with simple hyperplasia, 3% of patients with complex hyperplasia, 8% of patients with atypical simple hyperplasia, and 29% of patients with atypical complex hyperplasia. Most of the hyperplasias seemed to remain stable (18%) or regress (74%). As many as 25% to 43% of patients with atypical hyperplasia detected in an endometrial biopsy or curettage specimen will have an associated, usually welldifferentiated, endometrial carcinoma detected during hysterectomy (12).
Marked cytologic atypia, a high mitotic rate, and marked cellular stratification are features of atypical endometrial hyperplasia most often associated with the finding of an undiagnosed carcinoma at hysterectomy.
Fertility-Sparing Treatment of Endometrial Hyperplasia and Cancer
Younger patients with endometrial cancer tend to have disorders such as polycystic ovarian syndrome, chronic anovulation, and infertility, indicative of exposure to intrinsic estrogen excess (13). Lesions in this age group are usually well differentiated and of endometrioid subtype with the potential to regress with oral or localized progestational therapy. Although standard treatment for all endometrial cancer is hysterectomy and staging, nonsurgical treatment with hormonal therapy may be an option for appropriately selected women desiring to preserve fertility. Surrogate staging techniques, such as magnetic resonance imaging (MRI), may be employed to evaluate the depth of myometrial invasion or identify extrauterine disease, although low sensitivity has the potential for underdiagnosis (14,15).
High regression rates for endometrial cancer and atypical hyperplasia following treatment with progestin therapy are extensively documented (16–20). Relatively small cohorts of patients and reports of hormone failure suggest caution when counseling patients for conservative management (21). A comprehensive review of hormonal treatment of grade 1endometrial cancer, including 27 articles with a combined total of 81 patients, included treatment with a variety of progestational agents. The overall response rate was 76% (62/81) and the median time to regression was 12 weeks (22). The recurrence rate was 24% among responders; nearly all recurrences occurred within 1 year of diagnosis. Only 1 month of progestational treatment was required to achieve a response in the 76% of patients without recurrence. Twenty patients achieved pregnancy following treatment. It is important to note that 24% (19/81) of the original cohort never responded to treatment, and only 68% had any documented follow-up endometrial sampling.
Progestational therapy can successfully treat disease while preserving fertility for patients with atypical hyperplasia and well-differentiated presumed stage I endometrial cancer. Appropriate patient selection and exclusion criteria remain undefined. Patients must be counseled that failure to identify recurrence or extension of disease during progestational treatment may lead to a delay in definitive surgery and ultimately a compromised prognosis (18).
Continuous progestin therapy with megestrol acetate (40 to 160 mg per day) is probably the most reliable treatment for reversing complex or atypical hyperplasia, recognizing that higher doses may be associated with reduced compliance secondary to side effects. No clear consensus exists for an optimal follow-up interval. Therapy should be continued for at least 3 to 6 months, and endometrial biopsy should be performed to assess response. Periodic endometrial biopsy or transvaginal ultrasonography is advisable in patients being monitored on progestin therapy for atypical hyperplasia because of the presence of undiagnosed cancer in at least 25% of cases and the high recurrence rate after treatment with progestins. In this setting the use of progesterone should be considered a temporary, rather than long-term, treatment.
For women with atypical complex hyperplasia who no longer desire fertility, hysterectomy is recommended.
ENDOMETRIAL CANCER SCREENING IN THE GENERAL
POPULATION
Screening for endometrial cancer should not be undertaken because of the lack of an appropriate, cost-effective, and acceptable test that reduces mortality. Routine Papanicolaou (Pap) testing is inadequate, and endometrial cytologic assessment is too insensitive and nonspecific to be useful in screening for endometrial cancer, even in a high-risk population. Transvaginal ultrasonographic examination of the uterus and endometrial biopsy are too expensive to be employed as screening tests.
Screening for endometrial cancer or its precursors may be justified for certain high-risk women, such as members of families with hereditary nonpolyposis colorectal cancer (23). Women taking tamoxifen receive no benefit from routine screening with transvaginal ultrasonography or endometrial biopsy (24,25). Similarly, there is insufficient evidence to recommend screening for endometrial cancer in women because of a history of unopposed estrogen therapy, late menopause, nulliparity, infertility or failure to ovulate, obesity, diabetes, or hypertension (26).
FIGURE 37-1 Atypical hyperplasia (complex hyperplasia with severe nuclear atypia) of endometrium. A: The proliferative endometrial glands reveal considerable crowding and papillary infoldings. The endometrial stroma, although markedly diminished, can still be recognized between the glands. B: Higher magnification demonstrates disorderly nuclear arrangement and nuclear enlargement and irregularity. Some contain small nucleoli.
(Provided by Gordana Stevanovic, MD, and Jianyu Rao, MD, Department of Pathology,
UCLA.)
Most patients with endometrial cancer present with abnormal perimenopausal or postmenopausal uterine bleeding when the tumor is still confined to the uterus. Application of an appropriate and accurate diagnostic test in this situation usually results in early diagnosis and a high cure rate with timely treatment. It is important to recognize that the workup of abnormal uterine bleeding should include endometrial biopsy even in premenopausal patients, as 5% of cases are in women under the age of 40.
Surveillance and Prevention in Patients at High Risk
Most endometrial carcinomas are sporadic, but at least 10% of cases have a hereditary basis (27,28). Two genetic models were described in the development of familial endometrial cancer: Lynch II syndrome and patients with a predisposition for endometrial cancer alone, both of which are inherited in an autosomal dominant fashion (29). The majority of studies focused on the increased incidence of endometrial cancer associated with Lynch II syndrome, a highly penetrant disorder (80% to 85%) (30). Lynch II syndrome is caused by an inherited mutation in one of the following mismatch repair genes: hMSH2, hMLH1, PMS1, PMS2, or hMSH6 (31). The disorder is characterized by early age (average age younger than 45 years) at onset of neoplastic lesions in a variety of tissues, including the colon, uterus, stomach, ureters, ovaries, and skin (30,32).
The lifetime risk of endometrial cancer in women with Lynch II syndrome is 32% to 60% and the lifetime risk of ovarian cancer is 10% to 12% (33,34). Colorectal cancer is less prevalent in women with Lynch II syndrome than in men, whose risk approaches 100%. In a study of 1,763 patients from 50 Lynch II syndrome families in the Finnish Cancer Registry, the cumulative incidence of colorectal cancer in women was 54% by age 70, while the cumulative incidence of endometrial cancer was 60%, supporting surveillance (9,33,35). No effective screening method has been identified for patients at increased risk for ovarian cancer.
The Mallorca group, composed of experts in the treatment of hereditary gastrointestinal cancers from nine countries, recommended the following endometrial cancer surveillance strategy for patients with Lynch II syndrome: annual pelvic examination, transvaginal ultrasound, and endometrial biopsy beginning at 35 to 40 years of age (36). These recommendations are by expert opinion only, and it is unknown whether these interventions are cost-effective or will impact mortality from endometrial or ovarian cancer in patients with Lynch II syndrome. An attractive alternative to early detection is prophylactic surgery after completion of childbearing (37). A multi-institutional, matched case-control study found that prophylactic hysterectomy with bilateral salpingo-oophorectomy is an effective primary prevention strategy in women with Lynch II syndrome. No woman with hysterectomy and bilateral salpingo-oophorectomy developed endometrial, ovarian, or peritoneal carcinoma during the period of follow-up. In contrast, endometrial cancer developed in 33% and ovarian cancer in 5% of women who did not undergo prophylactic surgery (34).
Recognizing that 3% to 5% of endometrial cancers can be attributed to Lynch syndrome, a new diagnosis of endometrial cancer represents an opportunity to identify patients with germline mutations. Clinical screening using protocols such as the Amsterdam or Bethesda criteria has been used in the past, but the use of molecular tumor testing has been shown to improve the rate of Lynch II detection (38). For example, in 2007 the Society of Gynecologic Oncology developed clinical practice guidelines to identify women who may benefit from germline Lynch testing. Bruegl et al. attempted to validate these guidelines in an unselected cohort of patients with endometrial cancer and reported poor sensitivity (32.6%, 95% confidence interval [CI] 19.2%–48.5%), showing that a majority of Lynch II patients were missed using this criteria. They went on to show that universal screening was ultimately more cost effective than the SGO model (39). Another study by Goverde et al. found universal testing in endometrial cancer patients under age 70 was cost-effective (6,668 British pounds/life year gained) as a result of the prevention of colorectal cancer in index patients and their relatives (40).
Performing immunohistochemistry for MLH1, MSH2, MSH6, and PMS2 expression is one cost effective model that is available in most laboratories. Recognizing that loss of MLH1 is most often caused by hypermethylation, such analyses should be completed to appropriately triage patients to germline testing (41).
There are rare reports of pedigrees in which family members are affected by endometrial cancer alone, and genetic studies have not found a germline mutation associated with site-specific endometrial cancer (29,42,43). A population-based study of endometrial cancer and familial risk in younger women (Cancer and Steroid Hormone, or CASH, Study Group) reported that a history of endometrial cancer in a first-degree relative increased the risk of endometrial cancer by nearly threefold (odds ratio of 2.8; 95% CI, 1.9%–4.2%) (44). A significant association was found with colorectal cancers, with an observed odds ratio of 1.9 (95% CI, 1.1%–3.3%). The presence of Lynch II syndrome families within the cohort may explain the latter association, but a family history of endometrial cancer was an independent risk factor for endometrial cancer, after adjusting for age, obesity, and number of relatives (44).
Endometrial cancer and breast cancer share some of the same reproductive and hormonal risk factors such as nulliparity and exposure to unopposed estrogen (4,45,46). However, a familial association between breast and endometrial cancer is uncertain (46). Although it has been posited that patients with BRCA mutations were at elevated risk for endometrial cancer (in addition to breast and ovarian cancer), this increase in risk may be present only in those patients with a personal history of breast cancer who are taking tamoxifen (47).
ENDOMETRIAL CANCER
Clinical Features
Symptoms
Endometrial carcinoma most often occurs in women in the sixth and seventh decades of life, at an average age of 60 years; 75% of cases occur in women older than 50 years of age. About 90% of women with endometrial carcinoma have vaginal bleeding or discharge as their only presenting symptom. Most women recognize the significance of postmenopausal bleeding and seek medical consultation within 3 months. Some women experience pelvic pressure or discomfort indicative of uterine enlargement or extrauterine disease spread.
Bleeding may not have occurred because of cervical stenosis, especially in older patients, and may be associated with hematometra or pyometra, causing a purulent vaginal discharge. This finding is often associated with a poor prognosis (48). Less than 5% of women diagnosed with endometrial cancer are asymptomatic. In the absence of symptoms, endometrial cancer is usually detected following an abnormal Pap test, discovery of cancer in a uterus removed for some other reason, or evaluation of an abnormal finding on a pelvic ultrasonography examination or computed tomography (CT) scan obtained for an unrelated reason. Women who are found to have malignant cells on Pap test are more likely to have other poor prognostic factors (49).
Abnormal perimenopausal and postmenopausal bleeding should always be taken seriously and be properly investigated, no matter how minimal or nonpersistent. Causes may be nongenital, genital extrauterine, or uterine. Nongenital tract sites should be considered based on the history or examination, including testing for blood in the urine and stool. Invasive tumors of the cervix, vagina, and vulva are usually evident on examination, and any abnormalities should be biopsied. Traumatic bleeding from an atrophic vagina may account for up to 15% of all causes of postmenopausal vaginal bleeding. This diagnosis can be considered if inspection reveals a thin, friable vaginal wall, but the possibility of a uterine source of bleeding must first be eliminated. Possible uterine causes of perimenopausal or postmenopausal bleeding include endometrial atrophy, endometrial polyps, estrogen therapy, hyperplasia, and cancer (50,51) (Table 37-3). Uterine leiomyomas should never be accepted as a cause of postmenopausal bleeding. Endometrial atrophy is the most common endometrial finding in women with postmenopausal bleeding, accounting for 60% to 80% of such bleeding. Women with endometrial atrophy usually were menopausal for about 10 years. Endometrial biopsy often yields insufficient tissue or only blood and mucus, and usually bleeding ceases after biopsy.
Endometrial polyps account for 2% to 12% of postmenopausal bleeding. Polyps are often difficult to identify with office endometrial biopsy or curettage. Hysteroscopy, transvaginal ultrasonography, or both, may be useful adjuncts in identifying endometrial polyps. Unrecognized and untreated polyps may be a source of continued or recurrent bleeding, leading eventually to unnecessary hysterectomy.
Estrogen therapy is an established risk factor for endometrial hyperplasia and cancer. The risk for endometrial cancer is four to eight times greater in postmenopausal women receiving unopposed estrogen therapy, and the risk increases with time and higher estrogen doses. This risk can be decreased by the addition of a progestin to the estrogen, either cyclically or continuously.
Endometrial biopsy should be performed as indicated to assess unscheduled bleeding or annually in women not taking a progestin. Endometrial hyperplasia occurs in 5% to 10% of patients with postmenopausal uterine bleeding. The sources of excess estrogen should be considered, including obesity, exogenous estrogen, or an estrogen-secreting ovarian tumor. Less than 10% of patients with postmenopausal bleeding have endometrial cancer.
Table 37-3 Causes of Postmenopausal Uterine Bleeding
Cause of Bleeding Percentage
Endometrial atrophy 60–80
Estrogen replacement therapy 15–25
Endometrial polyps 2–12
Endometrial hyperplasia 5–10
Endometrial cancer 10
Premenopausal women with endometrial cancer invariably have abnormal uterine bleeding, which is often characterized as menometrorrhagia or oligomenorrhea, or cyclical bleeding that continues past the usual age of menopause. The diagnosis of endometrial cancer must be considered in premenopausal women if abnormal bleeding is persistent, recurrent or if obesity or chronic anovulation is present.
Signs
Physical examination seldom reveals any evidence of endometrial carcinoma, although obesity is a commonly associated constitutional factor. Special attention should be given to the more common sites of metastasis. Peripheral lymph nodes should be assessed carefully. Abdominal examination is usually unremarkable, except in advanced cases in which ascites or hepatic or omental metastases may be palpable. On gynecologic examination, the vaginal introitus and suburethral area, and the entire vagina and cervix, should be carefully inspected and palpated. Bimanual rectovaginal examination should be performed specifically to evaluate the uterus for size and mobility, the adnexa for masses, the parametria for induration, and the cul-de-sac for nodularity.
Diagnosis
[2] Office endometrial aspiration biopsy is the accepted first step in evaluating a patient with abnormal uterine bleeding or suspected endometrial pathology. The diagnostic accuracy of office-based endometrial biopsy is 90% to 98% when compared with subsequent findings at dilation and curettage (D&C) or hysterectomy (52).
The narrow plastic cannulas are relatively inexpensive, often can be used without a tenaculum, cause less uterine cramping (improving patient acceptance), and are successful in obtaining adequate tissue samples in more than 95% of cases. If cervical stenosis is encountered, a paracervical block can be performed, and the cervix can be dilated. Premedication with an antiprostaglandin agent can reduce uterine cramping. Complications following endometrial biopsy are exceedingly rare; uterine perforation occurs in only 1 to 2 cases per 1,000. Endocervical curettage may be performed at the time of endometrial biopsy if cervical pathology is suspected. A Pap test is an unreliable diagnostic test because only 30% to 50% of patients with endometrial cancer have abnormal Pap test results (53).
Hysteroscopy and D&C should be reserved for situations in which cervical stenosis or patient tolerance does not permit adequate evaluation by aspiration biopsy, bleeding recurs after a negative endometrial biopsy, or the specimen obtained is inadequate to explain the abnormal bleeding. Hysteroscopy is more accurate in identifying polyps and submucous myomas than endometrial biopsy or D&C alone (54).
Transvaginal ultrasonography may be a useful adjunct to endometrial biopsy for evaluating abnormal uterine bleeding and selecting patients for additional testing (55). Transvaginal ultrasonography, with or without endometrial fluid instillation (sonohysterography), may be helpful in distinguishing between patients with minimal endometrial tissue whose bleeding is related to perimenopausal anovulation or postmenopausal atrophy and patients with significant amounts of endometrial tissue or polyps who are in need of further evaluation. The finding of an endometrial thickness greater than 4 mm, a polypoid endometrial mass, or a collection of fluid within the uterus requires further evaluation. Although most studies agree that an endometrial thickness of less than 5 mm in a postmenopausal woman is consistent with atrophy, more data are needed before ultrasonography findings can be considered to eliminate the need for endometrial biopsy in a patient with symptoms (56).
Pathology
The histologic classification of carcinoma arising in the endometrium is shown in Table 37-4 (10,57).
Endometrioid Adenocarcinoma
The endometrioid type of adenocarcinoma accounts for about 80% of endometrial carcinomas. These tumors are composed of glands that resemble normal endometrial glands; they have columnar cells with basally oriented nuclei, little or no intracytoplasmic mucin, and smooth intraluminal surfaces (Fig. 37-2).
As tumors become less differentiated, they contain more solid areas, less glandular formation, and more cytologic atypia. The well-differentiated lesions may be difficult to separate from atypical hyperplasia.
Table 37-4 Classification of Endometrial Carcinomas
Endometrioid adenocarcinoma
Variants
Villoglandular or papillary
Secretory
With squamous differentiation
Mucinous carcinoma
Papillary serous carcinoma
Clear cell carcinoma
Squamous carcinoma
Undifferentiated carcinoma
Mixed carcinoma
FIGURE 37-2 Well-differentiated adenocarcinoma of endometrium. The glands and complex papillae are in direct contact with no intervening endometrial stroma, the socalled back-to-back pattern. (Provided by Gordana Stevanovic, MD, and Jianyu Rao, MD, Department of Pathology, UCLA.)
Criteria that indicate the presence of invasion and are used to diagnose carcinoma are desmoplastic stroma, back-to-back glands without intervening stoma, extensive papillary pattern, and squamous epithelial differentiation. These changes, with the exception of the infiltrating pattern with desmoplastic reaction, require an area of involvement equal to or exceeding one-half of a low-power microscopic field (LPF) (>1 LPF; 4.2 mm in diameter) (58,59).
The differentiation of a carcinoma, expressed as its grade, is determined by architectural growth pattern and nuclear features (Table 37-5). In the International Federation of Gynecology and Obstetrics (FIGO) grading system proposed in 1989, tumors are grouped into three grades: grade 1, 5% or less of the tumor shows a solid growth pattern; grade 2, 6% to 50% of the tumor shows a solid growth pattern; and grade 3, more than 50% of the tumor shows a solid growth pattern. The presence of notable nuclear atypia that is inappropriate for the architectural grade increases the tumor grade by one.
Adenocarcinomas with squamous differentiation are graded according to the nuclear grade of the glandular component. This FIGO system is applicable to all endometrioid carcinomas, including its variants, and to mucinous carcinomas. In serous and clear cell carcinomas, nuclear grading takes precedence; however, most investigators believe that these two carcinomas should always be considered high-grade lesions, making grading unnecessary.
Table 37-5 FIGO Definition for Grading of Endometrial Carcinoma
Histopathologic Degree of Differentiation:
G1 <5% nonsquamous or nonmorular growth pattern
G2 6–50% nonsquamous or nonmorular growth pattern
G3 >50% nonsquamous or nonmorular growth pattern
Notes on Pathologic Grading:
Notable nuclear atypia, inappropriate for the architectural grade, raises a grade 1 (G1) or
grade 2 (G2) tumor by one grade
In serous adenocarcinoma, clear cell adenocarcinoma, and squamous cell carcinoma,
2375nuclear grading takes precedence
Adenocarcinomas with squamous differentiation are graded according to the nuclear
grade of the glandular component
FIGO Committee on Gynecologic Oncology. Revised FIGO staging for carcinoma of the vulva, cervix, and endometrium. Int J Gynecol Obst 2009;105:103–104.
About 15% to 25% of endometrioid carcinomas have areas of squamous differentiation (Fig. 37-3). In the past, tumors with benign-appearing squamous areas were called adenoacanthomas, and tumors with malignant-looking squamous elements were called adenosquamous carcinomas. It is recommended that the term endometrial carcinoma with squamous differentiation be used to replace these two designations because the degree of differentiation of the squamous component parallels that of the glandular component, and the behavior of the tumor is largely dependent on the grade of the glandular component (60).
FIGURE 37-3 Adenocarcinoma with squamous differentiation of endometrium. This lesion is also classified as adenoacanthoma. Squamous cells with eosinophilic cytoplasm and distinct cell borders form solid clusters in the lumina of neoplastic glands. (Provided by Gordana Stevanovic, MD, and Jianyu Rao, MD, Department of Pathology, UCLA.)
A villoglandular configuration is present in about 2% of endometrioid carcinomas (61). In these tumors, the cells are arranged along fibrovascular stalks, giving a papillary appearance but maintaining the characteristics of endometrioid cells. The villoglandular variants of endometrioid carcinomas are always well-differentiated lesions that behave like the regular endometrioid carcinomas, and they should be distinguished from serous carcinomas. Secretory carcinoma is a rare variant of endometrioid carcinoma that accounts for about 1% of cases (62). It occurs mostly in women in their early postmenopausal years. The tumors are composed of well-differentiated glands with intracytoplasmic vacuoles similar to early secretory endometrium. These tumors behave as regular well-differentiated endometrioid carcinomas and have an excellent prognosis. Secretory carcinoma may be an endometrioid carcinoma that exhibits progestational changes, but a history of progestational therapy is rarely elicited. Secretory carcinoma must be differentiated from clear cell carcinoma because both tumors have predominately clear cells. These two tumors can be distinguished by their structure: secretory carcinomas have uniform glandular architecture, uniform cytology, and low nuclear grade, whereas clear cell carcinomas have more than one architectural pattern and a high nuclear grade.
Mucinous Carcinoma
About 5% of endometrial carcinomas have a predominant mucinous pattern in which more than one-half of the tumor is composed of cells with intracytoplasmic mucin (63). Most of these tumors have a well-differentiated glandular architecture; their behavior is similar to that of common endometrioid carcinomas, and the prognosis is good. It is important to recognize mucinous carcinoma of the endometrium as an entity and to differentiate it from endocervical adenocarcinoma. Features that favor a primary endometrial carcinoma are the merging of the tumor with areas of normal endometrial tissue, presence of foamy endometrial stromal cells, presence of squamous metaplasia, or presence of areas of typical endometrioid carcinoma.
Serous Carcinoma
About 3% to 4% of endometrial carcinomas resemble serous carcinoma of the ovary and fallopian tube (64). Most often, these tumors are composed of fibrovascular stalks lined by highly atypical cells with tufted stratification (Fig. 37-4). Psammoma bodies are frequently observed.
[3] Serous carcinomas, also referred to as uterine papillary serous carcinomas, are considered high-risk lesions. The first description in 1982 noted that this entity usually occurred in elderly, hypoestrogenic women who presented with advanced-stage disease and accounted for up to one-half of deaths from endometrial carcinoma (64). Since then, several reports documented the aggressive nature and poor prognosis of serous carcinomas.
FIGURE 37-4 Serous carcinoma of endometrium. Branching papillae are supported by delicate fibrovascular cores and lined with columnar cells with moderate nuclear atypism, multiple nucleoli, and mitotic figures. (Provided by Gordana Stevanovic, MD, and Jianyu Rao, MD, Department of Pathology, UCLA.)
They are commonly admixed with other histologic patterns, but mixed tumors behave as aggressively as pure serous carcinomas. Even patients with a very small proportion of serous features (5%) remain at high risk of recurrence (65). Serous carcinomas are often associated with lymph–vascular space and deep myometrial invasion. The presence of lymph node metastases, positive peritoneal cytology, and intraperitoneal tumor does not necessarily correlate with increasing myometrial invasion. Even when these tumors appear to be confined to the endometrium or endometrial polyps without myometrial or vascular invasion, they behave more aggressively than endometrioid carcinomas and have a propensity to spread intra-abdominally, simulating the behavior of ovarian carcinoma. In one series, 37% of patients with serous carcinomas of the endometrium confined to a polyp demonstrated extrauterine disease when subjected to exploration and surgical staging (66).
A multi-institutional review of 206 patients with surgical stage I and II serous carcinomas demonstrated recurrence in 21% (65). Substage and treatment with platinum-based chemotherapy were associated with improved overall survival. Survival of surgically staged patients without myometrial invasion or extrauterine disease is between 89% and 100%, suggesting that observation may be appropriate in select patients, particularly in elderly patients with comorbidities (67). However, stage I patients, particularly those with myometrial invasion, remain at high risk of peritoneal and vaginal recurrence. Therefore, platinumbased chemotherapy and vaginal brachytherapy should be considered in these patients (67–69).
Surgical treatment of advanced disease is no different from the endometrioid subtype, consisting of complete extirpation of visible disease (68). In one investigation from the Mayo Clinic, cytoreduction to microscopic residual was associated with a median overall survival of 51 versus 12 months for those patients with any residual (70). Postoperative treatment of advanced disease in the United States consists of chemotherapy and pelvic radiation, with or without paraaortic radiation. The Gynecologic Oncology Group (GOG) study GOG184 included serous carcinomas and randomized patients to carboplatin and paclitaxel versus cisplatin, doxorubicin (Adriamycin), and paclitaxel together with tumor volume–directed radiation (71). The former regimen demonstrated similar outcomes with less toxicity. Limited data suggest that delivering radiation “sandwiched” with chemotherapy improves progression-free and overall 3-year survival rates (72). Ongoing studies are evaluating the role of chemotherapy alone for these tumors, especially because of the high rate of peritoneal dissemination and recurrences. It remains unknown whether radiation improves survival when added to chemotherapy.
Clear Cell Carcinoma
Clear cell carcinoma accounts for less than 5% of all endometrial carcinomas (73). Clear cell carcinoma usually has a mixed histologic pattern, including papillary, tubulocystic, glandular, and solid types. The cells have highly atypical nuclei and abundant clear or eosinophilic cytoplasm. Often, the cells have a hobnail configuration arranged in papillae with hyalinized stalks (Fig. 37-5). Clear cell carcinoma characteristically occurs in older women and like serous carcinoma is considered a poor prognosticator. Traditionally clear cell carcinoma was associated with very poor outcomes with overall survival rates varying from 33% to 64%. A multi-institutional review of 99 patients with uterine clear cell carcinoma documented only one recurrence (vaginal) in the 22 patients without extrauterine disease subjected to thorough surgical staging (74). Considering all 49 patients with stage I or II disease (regardless of the extent of staging), only one hematologic failure was noted. These data argue against the use of systemic therapy in patients with clear cell carcinoma limited to the pelvis, while the 10% vaginal cuff failure suggests that vaginal brachytherapy alone may be sufficient treatment. In contrast, others argued for systemic treatment of patients with stage I disease (75).
Complete surgical staging is important because 52% of patients with clinical stage I clear cell carcinoma have metastatic disease. Patients who undergo a complete cytoreduction appear to have improved progression-free and overall survivals compared to women left with residual disease following surgery (74). Postoperative therapy for patients with advanced disease is platinum based (75).
Squamous Carcinoma
Squamous carcinoma of the endometrium is rare. Some tumors are pure, but most have a few glands. To establish primary origin within the endometrium, there must be no connection with or spread from cervical squamous epithelium. Squamous carcinoma often is associated with cervical stenosis, chronic inflammation, and pyometra at the time of diagnosis. This tumor has a poor prognosis, with an estimated 36% survival rate in patients with clinical stage I disease (76).
Simultaneous Tumors of the Endometrium and Ovary
Synchronous endometrial and ovarian cancers are the most frequent simultaneously occurring genital malignancies, with a reported incidence of 1.4% to 3.8% (77,78). Most commonly, the ovarian and endometrial tumor are well-differentiated endometrioid adenocarcinomas of low stage, resulting in an excellent prognosis. Patients often are premenopausal and present with abnormal uterine bleeding. The ovarian cancer usually is discovered as an incidental finding and is diagnosed at an earlier stage because of the symptomatic endometrial tumor, leading to a more favorable outcome. Up to 29% of patients with endometrioid ovarian adenocarcinomas have associated endometrial cancer.
If more poorly differentiated, nonendometrioid histologic subtypes are present, or if the uterine and ovarian tumors are histologically dissimilar, the prognosis is less favorable. Immunohistochemical studies, flow cytometry, and assessment of molecular DNA patterns to detect loss of heterozygosity (LOH) may be helpful in distinguishing between metastatic and independent tumors, but the differential diagnosis can usually be determined by conventional clinical and pathologic criteria.
Pretreatment Evaluation
After establishing the diagnosis of endometrial carcinoma, the next step is to evaluate the patient thoroughly to determine the best and safest approach to management of the disease. A complete history and physical examination are of utmost importance. Patients with endometrial carcinoma are often elderly and obese with a variety of medical problems, such as diabetes mellitus and hypertension, which complicate surgical management. Any abnormal symptoms, such as bladder or intestinal symptoms should be evaluated.
FIGURE 37-5 Clear cell adenocarcinoma of the endometrium. Back-to-back glands lined by polygonal to columnar cells with distinct cell membrane, abundant granular to clear cytoplasm, and variably sized nuclei (including binucleated and multinucleated forms) with prominent nucleoli (magnification ×400). (Provided by Gordana Stevanovic, MD, and Jianyu Rao, MD, Department of Pathology, UCLA.)
On physical examination, attention should be directed to enlarged or suspicious lymph nodes, including the inguinal area, abdominal masses, and possible areas of cancer spread within the pelvis. Evidence of distant metastasis or locally advanced disease in the pelvis, such as gross cervical involvement or parametrial spread, may alter the treatment approach.
Chest radiography should be performed to exclude pulmonary metastasis and to evaluate the cardiorespiratory status of the patient. Other routine preoperative studies should include electrocardiography, complete blood and platelet counts, and blood type and screen. Other preoperative or staging studies are neither required nor necessary for most patients with endometrial cancer. Studies such as cystoscopy, colonoscopy, intravenous pyelography, and barium enema are not indicated unless dictated by patient symptoms, physical findings, or other laboratory tests (79). CT scanning of the abdomen and pelvis may be considered in patients with type II uterine cancer to determine if minimally invasive surgery is appropriate. Stage IV disease is usually clinically evident based on patient symptomatology and clinical examination.
Ultrasonography and MRI can be used to assess myometrial invasion preoperatively with a fairly high degree of accuracy (80,81). This information may be of use in planning the surgical procedure with regard to whether lymph node sampling should be undertaken.
Serum CA125, an antigen that is elevated in the blood of 80% of patients with advanced epithelial ovarian cancers, is elevated in most patients with advanced or metastatic endometrial cancer. In one study, 78% of endometrial cancer patients with lymph node metastases had an elevated preoperative CA125 level (82). Preoperative measurement of serum CA125 may help determine the extent of surgical staging and, if elevated, may be useful as a tumor marker in assessing response to subsequent therapy (83,84).
Clinical Staging
Clinical staging, according to the 1971 FIGO system, should be performed only in patients who are deemed not to be surgical candidates because of their poor medical condition or the degree of disease spread (85). The current FIGO staging is surgical, which has supplanted the old clinical system. With improvements in preoperative and postoperative care, anesthesia administration, and surgical techniques, almost all patients are medically suitable for operative therapy. A small percentage of patients will not be candidates for surgical staging because of gross cervical involvement, parametrial spread, invasion of the bladder or rectum, or distant metastasis.
Surgical Staging
Widely accepted management of endometrial cancer consists of hysterectomy, [4] removal of remaining adnexal structures, and appropriate surgical staging in patients considered at risk for extrauterine disease (86,87). Surgical staging has been recommended for patients with endometrial cancer since 1988 (87). In spite of this general recommendation, the incorporation of a systematic pelvic and para-aortic lymphadenectomy in all patients is not universally accepted (88,89).
This recommendation became more controversial after the publication of two
large prospective randomized trials that failed to demonstrate improved outcomes
for patients who underwent pelvic lymphadenectomy (90,91). These two studies
show differences in their design: in the ASTEC trial all women with clinical stage
I were included without exclusion criteria, whereas the Italian study excluded
women with stage IA and IB grade 1 tumors, and nonendometrioid malignancies.
In the Italian study, systematic nodal dissection was performed, as opposed to
pelvic node sampling in the ASTEC trial (median number of lymph nodes
harvested 30 vs. 12, respectively). The studies share characteristics that could lead
to misinterpretation of their results. The percentage of nodal positivity is low in
both studies (13% and 9%), suggesting that regardless of differences in exclusion
criteria, low-risk cases were included in both studies, thus diluting possible (if
any) therapeutic benefit of lymphadenectomy. Neither study used the information
derived from lymphadenectomy to target postoperative treatment (i.e., to spare
patients with negative nodes from radiotherapy or to target postoperative
treatment to the metastatic areas), thus eliminating a potential benefit.
GOG33 demonstrated that patients with absent or superficial myometrial
invasion have a low probability of lymphatic metastases (92), and this risk
approaches zero for patients with endometrioid grade 1 or 2 disease,
superficial myometrial invasion, and tumor diameter less than 2 cm (93). An
observational study reported a significant survival benefit of para-aortic
lymphadenectomy in patients at intermediate or high risk of recurrence (based on
presence of histologic grade 3 or deep myometrial invasion, or lymphovascular
invasion, or evidence of spread outside of the uterine corpus), compared to
patients who had hysterectomy with pelvic lymphadenectomy but without paraaortic lymphadenectomy. This benefit was not observed in patients with low-risk
endometrial cancer (94). The Postoperative Radiation Therapy in Endometrial
Carcinoma (PORTEC) study identified patients with stage IC, grade 3
endometrial carcinoma as being at high risk of early distant spread and death
when treated with hysterectomy only (no staging), followed by pelvic external
beam radiation therapy. These patients had a 31% risk of distant recurrence (89).
Patients with the greatest potential to benefit from surgical staging are those with
risk factors such as histologic grade 3, deep myometrial invasion, or
lymphovascular invasion (see Surgical Treatment for a discussion of sentinel
lymph node biopsy).
In summary, surgical staging should (a) identify patients with disseminated
disease who are at high risk of recurrence; (b) target postoperative treatment; (c)
reduce the number of patients potentially requiring postoperative treatment when the provided information is used appropriately (avoiding the risk of morbidity
without reasonable benefit); and (d) possibly eradicate lymphatic disease. In spite
of these potential benefits in high-risk patients, prospective randomized data
demonstrating a survival advantage or reduction in overall morbidity resulting
from a potential reduction of adjuvant treatment are not available.
Table 37-6 Carcinoma of the Endometrium (2008)
Stage Ia Tumor confined to the corpus uteri
IAa No or less than half myometrial invasion
IBa Invasion equal to or more than half of the myometrium
Stage IIa Tumor invades cervical stroma, but does not extend beyond the uterusb
Stage IIIa Local and/or regional spread of the tumor
IIIAa Tumor invades the serosa of the corpus uteri and/or adnexaec
IIIBa Vaginal and/or parametrial involvementc
IIICa Metastases to pelvic and/or para-aortic lymph nodesc
IIIC1a Positive pelvic nodes
IIIC2a Positive para-aortic lymph nodes with or without positive pelvic lymph
nodes
Stage IVa Tumor invades bladder and/or bowel mucosa, and/or distant metastases
IVAa Tumor invasion of bladder and/or bowel mucosa
IVBa Distant metastases, including intra-abdominal metastases and/or inguinal
lymph nodes
aEither G1, G2, or G3.
b
Endocervical glandular involvement only should be considered as Stage I and no longer
as Stage II.
cPositive cytology has to be reported separately without changing the stage.
FIGO Committee on Gynecologic Oncology. Revised FIGO staging for carcinoma of the
vulva, cervix, and endometrium. Int J Gynecol Obst 2009;105:103–104.
The FIGO published the updated surgical staging system for endometrial cancer (Table 37-6) (95). In comparison with recommendations from 1988, the new system introduces the following changes: (a) former stages IA and IB are combined; (b) former stage IIA was eliminated so that only the presence of cervical stroma involvement is considered stage II disease; (c) peritoneal cytologic findings positive for endometrial cancer are no longer a criterion for disease upstaging (although FIGO still recommends the collection of peritoneal washing, recognizing the predictive value of positive cytologic findings when combined with other poor prognosis factors); and (d) stage IIIC was divided into IIIC1 and IIIC2 in accordance with the absence or presence of positive para-aortic nodes. The presence of parametrial disease is now formally recognized as stage IIIB disease.
Prognostic Variables
Although disease stage is the most significant variable affecting survival, [5] a number of other individual prognostic factors for disease recurrence or survival are known, including tumor grade, histopathology, depth of myometrial invasion, patient age, and surgical-pathologic evidence of extrauterine disease spread (Tables 37-7 and 37-8). Other factors, such as tumor size, peritoneal cytology, hormone receptor status, flow cytometric analysis, and oncogene perturbations, are implicated as having prognostic importance.
Table 37-7 Surgical-Pathologic Findings in Clinical Stage I Endometrial Cancer
Surgical-Pathologic Finding Percentage of Patients
Histology
Adenocarcinoma 80
Adenosquamous 16
Other (papillary serous, clear cell) 4
Grade
1 29
2 46
3 25
2387Myometrial invasion
None 14
Inner third 45
Middle third 19
Outer third 22
Lymph–vascular space invasion 15
Isthmic tumor 16
Adnexal involvement 5
Positive peritoneal cytology 12
Pelvic lymph node metastasis 9
Aortic lymph node metastasis 6
Other extrauterine metastasis 6
Modified from Creasman WT, Morrow CP, Bundy BN, et al. Surgical pathologic spread patterns of endometrial cancer. Cancer 1987;60:2037–2041, with permission.
Age
In general, younger women with endometrial cancer have a better prognosis than older women. Two reports observed no deaths related to disease in patients with endometrial cancer diagnosed before 50 years of age (96,97). Another series demonstrated a 60.9% 5-year survival rate for patients older than 70 years of age, compared with 92.1% survival rate for patients younger than 50 years of age (98). Decreased survival was associated with an increased risk for extrauterine spread (38% vs. 21%) and deep myometrial invasion (57% vs. 24%) for these two groups. [7] The GOG reported 5-year survival rates of 96.3% for patients 50 years of age or younger; 87.3% for patients 51 to 60 years; 78% for patients 61 to 70 years; 70.7% for patients 71 to 80 years; and 53.6% for patients older than 80 years (99).
Increased risk for recurrence in older patients was related to a higher incidence of grade 3 tumors or unfavorable histologic subtypes; however, age appears to be an independent prognostic variable. Increasing patient age appears to be independently associated with disease recurrence in endometrial cancer (100,101). In one study, the mean age at diagnosis of patients who had recurrence or died of disease was 68.6 years, compared with 60.3 years for patients without recurrence. For every 1 year increase in age, the estimated rate of recurrence increased 7%. None of the patients younger than 50 years of age developed recurrent cancer, compared with 12% of patients aged 50 to 75 years and 33% of patients older than 75 years (102).
Histologic Type
Nonendometrioid histologic subtypes account for about 10% of endometrial cancers and carry an increased risk for recurrence and distant spread (103– 105). In a retrospective review of 388 patients treated at the Mayo Clinic for endometrial cancer, 52 (13%) had an uncommon histologic subtype, including 20 adenosquamous, 14 serous, 11 clear cell, and 7 undifferentiated carcinomas. In contrast to the 92% survival rate among patients with endometrioid tumors, the overall survival for patients with one of these more aggressive subtypes was only 33%. At the time of surgical staging, 62% of the patients with an unfavorable histologic subtype had extrauterine spread of disease (103).
Histologic Grade
Histologic grade of the endometrial tumor is strongly associated with prognosis (86,92,102,106–108). In one study, recurrences developed in 7.7% of grade 1 tumors, 10.5% of grade 2 tumors, and 36.1% of grade 3 tumors. Patients with grade 3 tumors were in excess of five times more likely to have a recurrence than were patients with grades 1 and 2 tumors. The 5-year disease-free survival rates for patients with grades 1 and 2 tumors were 92% and 86%, respectively, compared with 64% for patients with grade 3 tumors (149). Another study reported similar results, noting recurrences in 9% of patients with grades 1 and 2 tumors compared with 39% of patients with grade 3 lesions (107). Increasing tumor anaplasia is associated with deep myometrial invasion, cervical extension, lymph node metastasis, and both local recurrence and distant metastasis.
Table 37-8 Prognostic Variables in Endometrial Carcinoma
Age
Histologic type
Histologic grade
Myometrial invasion
Lymph–vascular space invasion
Isthmus–cervix extension
Adnexal involvement
Lymph node metastasis
Intraperitoneal tumor
Tumor size
Peritoneal cytology
Hormone receptor status
DNA ploidy/proliferative index
Genetic/molecular tumor markers
Tumor Size
Tumor size is a significant prognostic factor for lymph node metastasis and survival in patients with endometrial cancer (93,109,110). One report determined tumor size in 142 patients with clinical stage I endometrial cancer and found lymph node metastasis in 4% of patients with tumors 2 cm or smaller, in 15% of patients with tumors larger than 2 cm, and in 35% of patients with tumors involving the entire uterine cavity (108). Tumor size better defined an intermediate-risk group for lymph nodes metastasis (i.e., patients with grade 2 tumors with less than 50% myometrial invasion). Overall, these patients had a 10% risk for lymph node metastasis, but there was no nodal metastasis associated with tumors 2 cm or smaller, compared with 18% when tumors were larger than 2 cm. Five-year survival rates were 98% for patients with tumors 2 cm or smaller, 84% for patients with tumors larger than 2 cm, and 64% for patients with tumors involving the whole uterine cavity (109).
Hormone Receptor Status
Estrogen receptor and progesterone receptor levels are prognostic indicators for endometrial cancer independent of grade in several studies (111–114). Patients whose tumors are positive for one or both receptors have longer survival times than patients whose carcinomas lack the corresponding receptors. Even patients with metastasis have an improved prognosis with receptor-positive tumors (112). Progesterone receptor levels appear to be stronger predictors of survival than estrogen receptor levels, and the higher the absolute level of the receptors, the better the prognosis.
DNA Ploidy and Proliferative Index
About two-thirds of endometrial adenocarcinomas have a diploid DNA content as determined by flow cytometric analysis (113,115–118). The proportion of nondiploid tumors increases with stage, lack of tumor differentiation, and depth of myometrial invasion. In several studies, DNA content was related to clinical course of the disease, with death rates reported to be higher in women whose tumors contained aneuploid populations of cells. The proliferative index is related to prognosis.
Myometrial Invasion
Because access to the lymphatic system increases as cancer invades into the outer one-half of the myometrium, increasing depth of invasion is associated with increasing likelihood of extrauterine spread and recurrence (92,107,119). The association of depth of myometrial invasion with extrauterine disease and lymph node metastases was reported (92). Of patients without demonstrable myometrial invasion, only 1% had pelvic lymph node metastasis, compared with patients with outer one-third myometrial invasion who had 25% pelvic and 17% aortic lymph node metastases. Deep myometrial invasion (>50% for all stages; ≥66% for stage I) is the strongest predictor of hematogenous recurrence (120). Survival decreases with increasing depth of myometrial invasion. In general, patients with noninvasive or superficially invasive tumors have an 80% to 90% 5-year survival rate, whereas those with deeply invasive tumors have a 60% survival rate. The most sensitive indicator of the effect of myometrial invasion on survival is distance from the tumor–myometrial junction to the uterine serosa. Patients with tumors that are less than 5 mm from the serosal surface are at much higher risk for recurrence and death than those with tumors greater than 5 mm from the serosal surface (121).
Lymph–Vascular Space Invasion
Lymph–vascular space invasion (LVSI) appears to be an independent risk factor for recurrence and death from all types of endometrial cancer (92,122,123). The overall incidence of LVSI in early endometrial cancer is about 15%, although it increases with increasing tumor grade and depth of myometrial invasion. One study reported LVSI in 2% of grade 1 tumors and 5% of superficially invasive tumors, compared with 42% of grade 3 tumors and 70% of deeply invasive tumors (124). LVSI was demonstrated to be a strong predictor of lymphatic dissemination and lymphatic recurrence (125). Another study reported deaths in 26.7% of patients with clinical stage I disease with LVSI, compared with 9.1% of those without LVSI (126). Likewise, an 83% 5-year survival rate was reported for patients without demonstrable LVSI, compared with a 64.5% survival rate for those in whom LVSI was present (181). Using multivariate analysis, LVSI correlated significantly with survival of patients with stage I endometrial adenocarcinomas in another report (123).
Isthmus and Cervix Extension
The location of the tumor within the uterus is important. Involvement of the uterine isthmus, cervix, or both, is associated with an increased risk for extrauterine disease, lymph node metastasis, and recurrence. Cervical stromal invasion was a strong predictor of lymphatic dissemination and lymphatic recurrence, especially for pelvic lymph nodes (125). One study reported that if the fundus of the uterus alone was involved with tumor, there was a 13% recurrence rate, whereas if the lower uterine segment or cervix was involved with occult tumor, there was a 44% recurrence rate (105). A subsequent GOG study found that tumor involvement of the isthmus or cervix without evidence of extrauterine disease was associated with a 16% recurrence rate and a relative risk of 1.6 (86). Patients with cervical involvement tended to have higher-grade, larger, and more deeply invasive tumors, undoubtedly contributing to the increased risk for recurrence.
Peritoneal Cytology
Several historical reports noted increased recurrence rates and decreased survival rates and, on this basis, recommended treatment for positive cytology (127,128). Most of the studies included patients with other evidence of extrauterine disease spread and were performed without appropriate multivariate analysis and with patients who were incompletely staged. The GOG study critically analyzed 1,180 clinical stages I and II endometrial cancer patients in whom appropriate surgical and pathologic staging was performed (86). Considering only the 697 patients for whom peritoneal cytology status and adequate follow-up were available, 25 (29%) of 86 patients with positive cytology developed recurrence, compared with 64 (10.5%) of 611 patients with negative cytology. They noted that 17 of the 25 recurrences in the positive cytology group were outside the peritoneal cavity.
In contrast to these reports, an equal number of studies found no significant relationship between malignant peritoneal cytology and an increased incidence of disease recurrence in the absence of other risk factors such as extrauterine disease (128–130). Patients with positive peritoneal cytology as the only site of extrauterine disease (i.e., no adnexal or uterine serosal invasion) and without poor prognosticators (i.e., myometrial invasion more than 50%, nonendometrioid histologic subtype, grade 3, lymph-vascular space invasion, cervical invasion) have a very favorable outcome with an absence of extra-abdominal recurrences (131). These patients have an associated 5-year survival of 98% to 100% even when not treated with adjuvant therapy (99,142). Patients with positive cytology in addition to poor prognostic factors demonstrate a high rate (47%) of distant extra-abdominal failure and may potentially benefit from systemic chemotherapy.
Positive peritoneal cytology seems to have an adverse effect on survival only if the endometrial cancer has spread to the adnexa, peritoneum, or lymph nodes, not if the disease is otherwise confined to the uterus (132). These considerations led to the omission of cytology as a factor impacting stage in the FIGO 2009 staging criteria. The following conclusions may be reached regarding the prognostic implications of positive peritoneal cytology:
1. Positive peritoneal cytology is associated with other known poor prognostic factors.
2. Positive peritoneal cytology in the absence of other evidence of extrauterine disease or poor prognostic factors has no significant effect on recurrence and survival.
3. Positive peritoneal cytology, when associated with other poor prognostic factors or extrauterine disease, increases the likelihood for distant and intra-abdominal disease recurrence and has a significant adverse effect on survival.
4. Use of several different therapeutic modalities has not resulted in any proven benefit to patients with endometrial cancer and positive peritoneal cytology.
Stage IIIA: Adnexal or Uterine Serosal Involvement
Most patients with stage IIIA disease have other poor prognostic factors that place them at high risk for recurrence. One series described treatment of all patients with serosal or adnexal invasion (or both) with whole-abdomen radiotherapy. Failures were observed outside the abdomen in 100% of patients with full-thickness myometrial invasion or uterine serosal invasion, and in 20% to 25% of cases in the presence of isolated adnexal invasion (86,133). These patients may benefit from postoperative systemic chemotherapy.
Lymph Node Metastasis
Lymph node metastasis is the most important prognostic factor in clinical early-stage endometrial cancer. Of patients with clinical stage I disease, about 10% will have pelvic and 6% will have para-aortic lymph node metastases. Patients with lymph node metastases have almost a sixfold higher likelihood of developing recurrent cancer than patients without lymph node metastases. One study reported a recurrence rate of 48% with positive lymph nodes, including 45% with positive pelvic nodes and 64% with positive aortic nodes, compared with 8% with negative nodes. The 5-year disease-free survival rate for patients with lymph node metastases was 54%, compared with 90% for patients without lymph node metastases (99). One series examined patients with lymph node metastases in addition to other extrauterine sites of disease (vagina, uterine serosa, positive peritoneal cytology, adnexal invasion). The recurrence rates were 67% (41% extranodal) for those with lymphatic dissemination versus 32% (5% extranodal) for those with other sites of extrauterine disease spread (131).
Intraperitoneal Metastases
Extrauterine metastasis, excluding peritoneal cytology and lymph node metastasis, occurs in about 4% to 6% of patients with clinical stage I endometrial cancer. Gross intraperitoneal spread is highly correlated with lymph node metastases; one study noted that 51% of patients with intraperitoneal tumor had positive lymph nodes, whereas only 7% of patients without gross peritoneal spread had positive nodes (92). Extrauterine spread other than lymph node metastasis is significantly associated with tumor recurrence. Another study found that 50% of patients with extrauterine disease developed recurrence, compared with 11% of patients without extrauterine disease, making recurrence almost five times more likely in patients with extrauterine disease spread. The 5-year diseasefree survival rate for patients with nonlymphatic extrauterine disease was 50%, compared with 88% in other patients (99). Predictors of peritoneal relapse include stage IV disease or stage II or III disease with two or more of the following risk factors: cervical invasion, positive peritoneal cytology, positive lymph nodes, and nonendometrioid histology (134).
Types I and II Endometrial Carcinoma: Molecular Aberrations
Based on their etiologic and pathologic features, sporadic endometrial cancer is classified into two subtypes (2,87). Type I (endometrioid histology) represents the majority of lesions (approximately 80%), which are mostly low grade, estrogen receptor positive, associated with hyperestrogenism, and arise from atypical complex hyperplasia (135). Hyperestrogenism may be attributed to obesity with peripheral conversion of androgens to estrogens, anovulation, or exposure to excessive exogenous estrogen (87,136–138). Obesity, polycystic ovarian syndrome, tamoxifen use, and unopposed estrogen use are all associated with increased risk of endometrial cancer. Other associated findings include late onset of menopause, nulliparity, diabetes mellitus, and hypertension. The molecular basis for the progression from hyperplasia to invasive endometrial carcinoma as a result of hyperestrogenism remains unknown because the involvement of only a minority of factors is reproducible (139). In contrast, type II endometrial cancer (serous, clear cell carcinoma) appears to be unrelated to high estrogen levels and often develops in nonobese women.
Type II cancers arise from its precursor, endometrial intraepithelial carcinoma (EIC) adjacent to an atrophic endometrium background in relatively older women (140). Distinct molecular changes are associated with these two subtypes. Common genetic changes in endometrioid endometrial cancer include mutations in PTEN (141–144), or β-catenin genes (145). In contrast, type II cancers frequently demonstrate alterations in HER2/neu, p53, p16, E-cadherin, and loss of LOH (146). These distinct molecular alterations underscore prognostic differences. Type I endometrial cancer is limited to the uterus in 70% of cases with a 5-year survival greater than 85%. Type II endometrial cancer displays a more aggressive clinical course and a poor prognosis; even in tumors with little or no myometrial invasion, more than one in three patients will have extensive extrauterine spread with complete surgical staging, resulting in an overall survival of 20% (66,70,73,147).
Inactivation of the PTEN tumor–suppressor gene is one of the earliest aberrations observed in endometrial cancer precursors and is the most common genetic defect in type I cancers, observed in up to 83% of tumors (141). Tumors with PTEN mutations tend to be well differentiated and minimally invasive (148). Approximately 20% of sporadic endometrioid cancers demonstrate a molecular phenotype referred to as microsatellite instability (MSI) (149,150). Microsatellites are short segments of repetitive DNA bases scattered throughout the genome. MSI describes the accumulation of sequence changes in these DNA segments that occur because of the inactivation of intranuclear proteins that comprise the mismatch repair system (151). Inactivation of MLH1, a component of the mismatch repair system, is a common event in type I endometrial cancer. This alteration occurs through hypermethylation of CpG islands in the gene promoter, a process known as epigenetic silencing (152,153). This is in contrast to colon cancer, in which MSI and inactivation of the mismatch repair genes occurs through mutations in mismatch repair genes, including hMSH2, hMLH1, PMS1, PMS2, or hMSH6 (31). MSI and abnormal methylation of MLH1 are early events in endometrial carcinogenesis and are described in precancerous lesions (154,155). Mutations in codons 12 or 13 of the K-ras oncogene are reported in 10% to 20% of endometrial adenocarcinomas (156). The presence of mutations of K-ras appears to be an independent unfavorable prognostic factor (157,158).
Chromosomal instability with extensive genomic derangements is commonly found in type II endometrial cancers (159). The most frequent genetic alteration is TP53 mutation, present in about 90% of serous carcinomas (160,161). In contrast to endometrioid carcinoma, MSI is rare (<5%), as are K-ras and PTEN mutations (144,162). Other genetic alterations that occur more frequently in serous compared to endometrioid carcinomas are inactivation of p16 (45%) and overexpression and gene amplification of HER2/neu oncogene (45% and 70%, respectively) (160,163). HER-2/neu overexpression is related to diminished, progression-free survival (164,165). E-cadherin, an oncogene responsible for cellto-cell adhesion that seems to play a critical role in initiation and progression of endometrial neoplasia, is absent or reduced in 62% to 87% of cases. Loss of Ecadherin is often associated with advanced stage and LOH in both serous and clear cell carcinomas (166,167).
Surgical Treatment
The most common current protocol for surgical management of endometrial cancer includes peritoneal cytology, hysterectomy and bilateral salpingooophorectomy, and surgical staging. In patients with nonendometrioid cancer, omentectomy along with appendectomy and peritoneal biopsies may be performed. The need to perform lymphadenectomy continues to evolve from complete pelvic and para-aortic lymphadenectomy historically, to lymphadenectomy based on risk factors for dissemination, to the increasingly common use of sentinel lymph node biopsy (168,169). The National Comprehensive Cancer Network (NCCN)® states that lymph node assessment can be performed using either strategy (170).
Selective lymphadenectomy based on a surgical algorithm is determined by histology, tumor grade, tumor size, extent of myometrial invasion, and presence of extra-uterine disease. Lymphadenectomy is omitted altogether for patients without endometrioid tumors measuring less than 2 cm in size and less than 50% myometrial invasion. The decision to administer postoperative radiation, chemotherapy, or both, is predicated on the final results of pathologic examination of the surgical specimen and cytology, according to previously described criteria (168).
Sentinel lymphadenectomy is performed using a standardized cervical injection of indocyanine green and fluorescence imaging in patients with clinical stage I endometrial cancer of all histologies and grades. Identified sentinel lymph nodes are evaluated using hematoxylin, eosin, and immunohistochemistry staining. A multicenter trial showed a 97.2% sensitivity and negative predictive value of 99.6% for the sentinel lymph node technique (171). Patients with macrometastases in sentinel lymph nodes are treated according to previous algorithms (172). Adjuvant therapy for low-volume sentinel lymph node metastases is still under debate (173).
Vaginal Hysterectomy
Vaginal hysterectomy may be considered for patients with endometrial cancer and multiple morbidities and in some circumstances may be preferable to radiation therapy alone. Vaginal hysterectomy with bilateral salpingo-oophorectomy alone may be acceptable treatment for patients with low-risk tumors (endometrioid, grade 1 or 2, <50% myometrial invasion, and tumor diameter <2 cm). In one report, a 94% survival rate was found in 56 patients with clinical stage I endometrial carcinoma treated by vaginal hysterectomy, with or without postoperative radiotherapy (mostly brachytherapy). Three-fourths of these patients had grade 1 lesions (174). Other groups have reported excellent outcomes for patients treated with vaginal hysterectomy with curative intent (175,176).
Laparoscopic Management
Advances in endoscopic technologies and power sources allowed application of a laparoscopic approach to the management of endometrial cancer. Since 1992, there were multiple reports documenting the feasibility of laparoscopically assisted vaginal hysterectomy with bilateral salpingo-oophorectomy and laparoscopic lymphadenectomy for staging and treatment of patients with endometrial cancer (177–180). These early studies demonstrated no differences in lymph node counts, estimated blood loss, and recurrence or survival rates with laparoscopy versus laparotomy, whereas decreased perioperative morbidity, longer operating times, shorter hospital stays, and earlier return to work were associated with laparoscopy.
Although the literature has isolated case reports of port site metastases, there are few data documenting the incidence. Martinez et al. analyzed 1,216 patients with endometrial and cervical cancer and showed that the port site metastasis rate was less than 0.5%; no port site metastases occurred after excluding patients with peritoneal disease (181).
A large prospective study by GOG randomized patients to laparoscopy versus laparotomy for the primary treatment of patients with endometrial cancer (182). Over 2,500 patients were enrolled, 1,696 to laparoscopy and 920 to laparotomy. Consistent with early reports, patients randomized to laparoscopy had shorter hospital stay (52% more than 2 days vs. 94% in the laparotomy group), less blood loss, and fewer postoperative complications (14% vs. 21%). The rate of intraoperative complications was similar, and the operative time was longer in the laparoscopy cohort. There was no difference in lymph node counts, and stage distribution was identical between groups. A follow-up quality-of-life investigation of the same cohorts revealed improved Functional Assessment of Cancer Therapy–General (FACT-G) scores, better physical functioning, better body image, less pain and its interference with quality of life, and an earlier resumption of normal activities and return to work over the 6-week recovery period in the laparoscopic group (183). Although the differences were modest and negligible by 6 months, the analysis was performed as per intention to treat. Of concern was the 24% rate of conversions in the laparoscopic cohort; only 4% were converted because of advanced disease. Improvements in laparoscopic equipment and surgical technique have drastically decreased this conversion rate to 2.4% in a subsequent trial (184).
The use of extraperitoneal laparoscopic staging for para-aortic lymphadenectomy was reported in a prospective investigation of 293 patients with endometrial cancer (185). The extraperitoneal approach greatly improved exposure by circumventing handling of the small bowel and was successful in over 90% of unselected patients up to a BMI of 51. There was no difference in lymph node counts, all dissections were performed to the level of the renal veins, and significant improvements in blood loss, postoperative complications, and length of stay were noted compared to the laparotomy group. If full para-aortic lymphadenectomy is necessary in an obese patient, the extraperitoneal approach offers the most reliable method to consistently perform a thorough dissection. Robotic-assisted surgery gained popularity for endometrial cancer treatment (see Chapter 28). Improved instrumentation and visualization allow minimally invasive surgery to be performed by surgeons with less laparoscopic experience, and in patients, particularly obese patients, who otherwise might not be candidates for minimally invasive surgery. In one report limited to obese patients with endometrial carcinoma, the robotic approach offered reduced operating times, less blood loss, higher lymph node yields, and shorter hospital stays compared to laparoscopy (186).
When not limited to obese patients, robotic-assisted surgery appears to offer the same benefits as laparoscopy in regards to postoperative morbidity and convalescence compared to laparotomy, but with shorter operative times (187). Long-term survival in patients treated with laparoscopy for their endometrial cancer show comparable outcomes, and can be expected to apply to robotic surgery, as it is simply an advanced laparoscopic tool. A GOG trial showed threeyear recurrence rates of 11.4% with laparoscopy versus 10.2% with open surgery, and 5-year survival was 90% in both arms (188).
Radical Hysterectomy
Radical hysterectomy, with removal of the parametria and upper vagina, and bilateral pelvic lymphadenectomy, does not improve survival of patients with clinical stage I disease compared with extrafascial hysterectomy and bilateral salpingo-oophorectomy alone (189–191). Radical hysterectomy increases intraoperative and postoperative morbidity and should not be performed for treatment of apparent early endometrial cancer. In the presence of demonstrable invasion of the cervix, a modified extrafascial hysterectomy may be performed. This may improve outcomes and decrease the risk of local recurrences, especially if postoperative local radiation is not planned in younger patients (192).
Radiation Therapy as Primary Treatment
Primary surgery followed by individualized radiation therapy is the most widely accepted treatment for early-stage endometrial cancers. About 5% to 15% of endometrial cancer patients have severe medical conditions that render them unsuitable for surgery (86). These patients tend to be elderly and obese with multiple chronic or acute medical illnesses, such as hypertension, cardiac disease, diabetes mellitus, and pulmonary, renal, and neurologic diseases.
Several series show that radiotherapy is effective treatment for patients with inoperable endometrial cancer (Table 37-9) (193,194). One reported on the treatment of 120 patients with clinical stage I and 17 patients with clinical stage II endometrial cancer with radiation alone, 85% of whom received only intracavitary radiation. Because of the high incidence of death caused by intercurrent illness in this group of patients, the 5- and 10-year overall survival rates were only 55% and 28%, respectively, compared with disease-specific survival rates of 87% and 85%, respectively. There was no difference in diseasespecific survival rates between patients with stage I and II disease. Intrauterine cancer recurred in 14% of patients, and extrauterine pelvic disease recurred in 3%. The authors treated 15 patients with stage III and IV disease, usually with a combination of external-beam and intracavitary radiation therapy, yielding a 5- year disease-specific survival rate of 49%. Five patients (3%) had serious late complications of radiation therapy (194).
Although it is generally agreed that intracavitary radiation is necessary to achieve adequate local control, the indications for external-beam radiation therapy in the primary treatment of endometrial cancer are less well defined. Patients with cervical involvement and known or suspected extrauterine pelvic spread undoubtedly would benefit from external-beam radiation therapy. Theoretically, external-beam radiation could sterilize microscopic nodal disease and possibly increase the radiation dose to deep myometrial or subserosal uterine disease, which may receive an insufficient dose from intracavitary radiation alone. A correlation between tumor grade and recurrence was noted in several reports. One found that the 5-year progression-free survival rate for medically inoperable patients with clinical stage I disease treated with radiotherapy alone was 94% for grade 1, 92% for grade 2, and 78% for grade 3 tumors (193).
The decision to treat a patient who has endometrial cancer with radiation alone must include a careful analysis of the relative risks and benefits of surgery. Although radiation alone can produce excellent survival and local control, it should be considered for definitive treatment only if the operative mortality rate is estimated to exceed the 10% to 15% risk for uterine recurrence that is expected with radiation treatment alone.
Table 37-9 Review of Recent Series of Endometrial Carcinoma Treated with Radiation Alone
Patterns of Metastatic Dissemination: Implications for Postoperative and Disease-Based Adjuvant Treatment
Endometrial cancer is commonly diagnosed early in its natural history with approximately 80% of patients presenting with stage I disease. Nevertheless, approximately one of every three women who die of endometrial cancer was considered to have early locoregional disease at primary diagnosis. The majority of treatment failures and the accompanying compromised longevity probably result from the failure to recognize sites of occult extrauterine dissemination at primary diagnosis. Traditional postoperative therapy (modality based) for highrisk endometrial cancer is external-beam radiotherapy that is frequently supplemented with vaginal brachytherapy. This approach improves local control but not survival in early-stage disease (88,195).
Understanding the different pathways of metastatic dissemination of endometrial cancer and their predictive factors allows the development of an individualized model for target-based therapeutic approaches to the predicted site(s) of failure. The natural history of epithelial corpus cancer includes four potential routes of metastasis: (a) contiguous extension (mainly to the vagina); (b) hematogenous dissemination; (c) lymphatic embolization; and (d) exfoliation with intraperitoneal spread. On the basis of regression analysis, independent pathologic risk factors predictive of the four routes of metastatic spread were identified:
1. Contiguous extension: histologic grade 3 and lymph-vascular space invasion are proven predictors of vaginal relapse in stage I endometrial cancer (196).
2. Hematogenous: deep myometrial invasion is the strongest predictor of hematogenous recurrence (>50% for all stages and ê66% for stage I) (119,120).
3. Lymphatic: lymphatic failure is more likely to occur when cervical stroma involvement or positive lymph nodes are present (125).
4. Peritoneal: predictors of peritoneal relapse are: (a) stage IV disease; or (b) stage II or III disease with two or more of the following risk factors: cervical invasion, peritoneal cytologic results positive for endometrial cancer, positive lymph nodes, and nonendometrioid histologic findings (134).
Patients with the risk factors summarized in Table 37-10 account for 35% of the overall population with endometrial cancer, but 89% of the observed hematogenous, lymphatic, and peritoneal relapses. Importantly, 46% of the patients considered at risk subsequently experienced a recurrence in one or more of the three sites, compared with only 2% of patients not judged to be at risk based on these criteria (p <0.001). The identification of subgroups of patients at risk for the different patterns of recurrence would allow postoperative treatment targeted to the predicted areas of tumor dissemination. The recurrence sites predicted by risk factors would presuppose different adjuvant treatment strategies. Patients at risk for hematogenous or peritoneal recurrence would potentially benefit from systemic cytotoxic treatment, while patients at risk for lymphatic or vaginal recurrence would potentially benefit from radiation treatment directed at areas at risk.
Table 37-10 Rates of Recurrence at 5 Years According to the Different Risk
Categories for 915 Patients
Risk Category Recurrence at 5 Years (%)
Hematogenous
All stages
Myometrial invasion ≤50% 4
Myometrial invasion >50% 28
Stage I (negative lymph nodes)
Myometrial invasion <66% 2
Myometrial invasion ≥66% 34
Lymphatic
No risk factors 2
CSI and/or positive lymph nodes 31
Peritoneal
Stage IV disease 63
Stage II–III disease and ≥2 risk factorsa 21
Stage I–III disease and ≤1 risk factora 1
Overallb
Not at riskc 2
At riskc 46
aCSI, nonendometrioid histologic subtype, positive lymph nodes, or positive result on
peritoneal cytologic evaluation.
b
Excluding vaginal recurrences.
cFor at least one of the three categories of recurrence (i.e., hematogenous, lymphatic, or
peritoneal).
CSI, cervical stromal invasion.
From Mariani A, Dowdy SC, Keeney GL, et al. High-risk endometrial cancer subgroups:
Candidates for target-based adjuvant therapy. Gynecol Oncol 2004;95:120–126.
MODALITIES OF POSTOPERATIVE TREATMENT
Observation
Patients with grades 1 and 2 lesions without myometrial invasion or any of the above risk factors (Table 37-10) have an excellent prognosis and require no postoperative therapy. In a GOG study, there were no recurrences and a 100% disease-free 5-year survival rate in the 91 patients in this category, 72 of whom received no additional treatment after hysterectomy (86). Other investigators reported equally favorable results with only surgical therapy in similar patients (197).
Vaginal Vault Radiation
Vaginal brachytherapy is an attractive alternative to external radiation therapy (ERT). High-dose rate (HDR) brachytherapy is well tolerated with low rates of severe or chronic complications. Vaginal control rates with the more convenient, better-tolerated HDR brachytherapy are comparable to control rates with the lengthier low-dose rate (LDR) brachytherapy. Pearcey and Petereit established the HDR dosing of 21 Gy to 5-mm depth in three fractions as the standard brachytherapy dose, providing local control rates of 98% to 100% (198).
Additional retrospective data demonstrate 96% to 98% vaginal control rates with HDR in high-risk early-stage endometrial cancer (199,200). Retrospective data suggest that the vaginal relapse rate after brachytherapy averages 4% to 5% (Table 37-11) and this is similar to the 5-year vaginal failure rate of 3.5% reported among the highest-risk patients who received ERT in PORTEC-1 (88,199,201,202).
Table 37-11 Recurrence in High-Risk, Comprehensively Staged, Early-Stage Endometrial Cancer after Adjuvant Vaginal Brachytherapy Alone (No Pelvic ERT)
PORTEC-2 randomized patients with apparent uterine-confined endometrial cancer at high risk for recurrence (>60 years of age with grade 1 or 2, stage IB and grade 3, stage IA; or any age, any grade IIA with <50% myometrial invasion), to pelvic ERT (46 Gy, in 23 fractions) versus vaginal brachytherapy (21 Gy in three HDR fractions or 30 Gy LDR, to a depth of 0.5 cm). At 3 years, there was no difference in vaginal failure rates (0.9% for vaginal brachytherapy, 2% for pelvic ERT; p = 0.97). There was a higher rate of nonvaginal pelvic relapse in the brachytherapy group (3.6%) compared to the ERT group (0.7%; p = 0.03), however, the absolute difference was small and there was no difference in overall survival (203). The difference between nonvaginal pelvic recurrences may be a reflection of unrecognized lymph node metastases at the time of initial surgery treated with ERT. One concern regarding PORTEC-2 is that there was not a surgery-only control in the study. However, the highest-risk endometrial cancer subgroup in PORTEC-1 (patients >60 years of age with grade 3 or deeply invasive grade 1 or 2, all stage I) was similar to the cohort included in PORTEC-2 and the locoregional recurrence rate in patients who did not receive adjuvant ERT in PORTEC-1 was 18% (88).
Grade 3 histology and lymph-vascular invasion are proven predictors of vaginal relapse in stage I endometrial cancer. Patients with these risk factors are the most likely group to benefit from vaginal vault brachytherapy (196). Although vaginal recurrences can be successfully treated and controlled in up to 81% of cases, the addition of vaginal brachytherapy to the initial surgical intervention can significantly reduce the risk of such recurrences (204).
External Pelvic Radiation
Radiation therapy traditionally was suggested to patients who were deemed to have intermediate or high risk of recurrence, according to grade and depth of myometrial invasion. Several retrospective studies and large, randomized trials did not show an overall survival benefit for intermediate- and high-risk patients with stage I endometrial cancer (or occult IIA endometrial cancer according to the 1988 FIGO staging) who received adjuvant pelvic radiotherapy.
[6] The PORTEC trial tested the role of postoperative pelvic radiation therapy for presumed stage I endometrial cancer in 714 patients. Eligibility criteria were stage IB, grades 1 to 2, and stage IA, grades 2 to 3; patients with stage IA, grade 3, were only 10% of the study population, and lymph node biopsies and peritoneal cytology were not required. Local–regional recurrences developed in 14% of the surgery group, compared with 4% of the postoperative pelvic radiation group. Overall, the 5-year survival rate was no different between the two groups (85% vs. 81%, respectively) (88). These results were confirmed by GOG99, a prospective, randomized investigation of surgery alone (including lymphadenectomy to the level of the inferior mesentery artery in some patients) versus surgery plus adjuvant pelvic radiation in intermediate-risk endometrial cancer (stages IA to IIB occult). Of 392 patients accrued to the study, more than 80% were actually low-risk patients (90.6% stage I, 81.6% grades 1 to 2, 82% <50% myometrial invasion). Disease recurrence was reduced by 58% (p = 0.007) with the use of postoperative pelvic radiation. After 2 years, the cumulative recurrence rate was 12% in the group with no postoperative treatment compared with 3% in the group that received pelvic radiation. The pelvic failure rate was 8.9% in the surgery-alone group compared with 1.6% in the postoperative pelvic radiation group. Overall survival rates were not significantly improved in patients receiving postoperative pelvic radiation compared with those treated only with surgery (92% vs. 86%, respectively) (259). The intergroup ASTEC/EN.5 trial provided further confirmation that external-beam radiation therapy in patients at intermediate to high risk of recurrence has no significant effect on overall survival (205).
Postoperative whole-pelvis external-beam radiation usually involves the delivery of 4,500 to 5,040 cGy in 180 cGy daily fractions over 5 to 6 weeks to a field encompassing the upper one-half of the vagina inferiorly, the lower border of the L4 vertebral body superiorly, and 1 cm lateral to the margins of the bony pelvis. The dose of radiation at the surface of the vaginal apex usually is boosted to 6,000 to 7,000 cGy by a variety of techniques. The most frequently reported side effects are gastrointestinal, usually abdominal cramps and diarrhea, although more serious complications such as bleeding, proctitis, bowel obstruction, and fistula can occur and may require surgical correction. The urinary system may be affected in the form of hematuria, cystitis, or fistula. The overall complication rate ranges from 25% to 40%; and the rate of serious complications requiring surgical intervention is about 1.5% to 3%.
External-beam pelvic radiation does not appear to impact survival in patients with high-risk stage I endometrial cancer. Patients with extrauterine pelvic disease, including adnexal spread, parametrial involvement, and pelvic lymph node metastases, in the absence of extrapelvic disease, are likely to benefit from postoperative pelvic radiation.
Extended-Field Radiation
Patients with histologically proven para-aortic node metastases or patients at risk for para-aortic disease, but without surgical evaluation, are candidates for extended-field radiation therapy. The entire pelvis, common iliac lymph nodes, and para-aortic lymph nodes are included within the radiation field. The paraaortic radiation dose is limited to 4,500 to 5,000 cGy. Extended-field radiotherapy appears to improve survival in patients with endometrial cancer who have positive para-aortic lymph nodes (86,206).
Five-year survival rates of 47% and 43% were reported for patients with surgically confirmed isolated para-aortic lymph node metastases and for those with para-aortic and pelvic lymph node metastases, respectively, using postoperative extended-field radiation. In one report, only one case of severe enteric morbidity occurred in 48 patients, a complication rate of 2% (207). In a GOG study, 37 of 48 patients with positive para-aortic nodes received postoperative para-aortic radiation, 36% of whom remained tumor free at 5 years (86). A comparison of patients with positive para-aortic nodes treated with megestrol acetate alone versus megestrol acetate and extended-field radiation showed that the survival rate in the patients receiving extended-field radiation was significantly better: 53% versus 12.5%, respectively (208). In another study of 18 patients with positive para-aortic nodes, 5-year survival rates were 67% for microscopic nodal disease and 17% for gross nodal disease (209).
Whole-Abdomen Radiation
Whole-abdomen radiation therapy (WART) was historically reserved for patients with stages III and IV endometrial cancer. It was also considered for patients with serous or carcinosarcomas, which have a propensity for upper abdominal recurrence (210–213). WART is rarely used since the publication of GOG122, which showed favorable progression free and overall survival in advanced endometrial carcinoma with chemotherapy as opposed to WART (see section on “Chemotherapy”) (214). The recommended dose to the whole abdomen is 3,000 cGy in 20 daily fractions of 150 cGy, with kidney shielding at 1,500 to 2,000 cGy, along with an additional 1,500 cGy to the para-aortic lymph nodes and 2,000 cGy to the pelvis. Gastrointestinal side effects, including nausea, vomiting, and diarrhea, sometimes make it necessary to interrupt therapy, but it is rare for patients to discontinue treatment because of these symptoms.
Hematologic toxicity can be expected to occur during whole-abdomen radiation, but it is usually mild. The incidence of late complications, mainly chronic diarrhea and small bowel obstruction, is low (5% to 10%).
Progestins
Because most endometrial cancers have estrogen and progesterone receptors, and progestins were used successfully to treat metastatic endometrial cancer, postoperative adjuvant progestin therapy attempted to reduce the risk of recurrence. This therapy is attractive because it provides systemic treatment and has few side effects. Unfortunately, several large randomized, placebocontrolled studies failed to identify a benefit for adjuvant progestin therapy (215,216).
Chemotherapy
Adjuvant cytotoxic chemotherapy was studied in a few trials. The GOG treated 181 patients who had poor prognostic factors with postoperative radiation and randomly assigned patients to receive no further therapy or doxorubicin chemotherapy. After 5 years of observation, there was no difference in recurrence rates between the two groups (217).
The GOG122 trial compared whole-abdominal radiotherapy versus systemic chemotherapy (eight cycles of doxorubicin and cisplatin) in 388 patients with stage III or IV disease who underwent maximal surgical resection of disease to less than 2 cm. Its results showed a significant advantage of chemotherapy on 5- year survival (214). Patients who received chemotherapy had a 13% improvement in 2-year progression-free survival (50% vs. 46%) and an 11% improvement in overall 2-year survival (70% vs. 59%) compared with patients treated with wholeabdomen radiation. This study was the first to suggest an improvement in outcome for use of adjuvant chemotherapy compared with radiation, but toxicity was more prevalent with chemotherapy. Patients with gross residual disease were assigned to the radiation arm, almost guaranteeing failure; and overall, 55% of patients experienced a recurrence or progression during the study period (218).
GOG184 randomized 552 patients with advanced disease to six cycles of cisplatin and doxorubicin with or without paclitaxel following surgical debulking and radiotherapy. Side effects were more pronounced with the three-drug regimen, and recurrence-free survival (RFS) at 36 months was no different between arms (62% vs. 64% for the three-drug regimen). The investigation was closed to patients with stage IV disease during the trial, but subgroup analysis suggested a 50% reduction in recurrence or death in the 57 patients with gross residual disease who received cisplatin, doxorubicin, and paclitaxel (71).
The PORTEC-3 trial and GOG258 investigated the combination of chemotherapy and radiotherapy in advanced or high-risk endometrial cancer. PORTEC-3 is investigated overall survival and failure-free survival of patients with high-risk and advanced stage endometrial carcinoma treated after surgery with chemoradiation followed by chemotherapy versus pelvic radiation alone (219). The results of the large randomized trial showed that the addition of 4 cycles carboplatin and paclitaxel chemotherapy to standard external-beam radiotherapy resulted in a statistically significant improvement of progression-free survival (75.5% vs. 68.6%, HR = 0.71) with no significant benefit to overall survival (81.8% vs. 76.7%, HR = 0.76). The investigators recommended that women with high-risk endometrial cancer should be individually counseled about this combined treatment, and that continued follow-up is needed to evaluate long-term survival.
GOG258 compared chemoradiation followed by chemotherapy versus chemotherapy alone in advanced endometrial cancer (220). The open-label randomized phase III trial compared treatment with cisplatin and tumor volumedirected radiation followed by carboplatin and paclitaxel for four cycles (C-RT) versus carboplatin and paclitaxel treatment for six cycles (CT). The preliminary analysis showed that combined modality regimen did not increase RFS in patients with optimally debulked stage III/IVA endometrial carcinoma (HR 0.9, 95% CI [0.74,1.1]). C-RT reduced the rate of local recurrence compared to CT. The estimates for 5-year overall survival are 70% for the C-RT arm versus 73% for the CT arm, and long-term survival and quality-of-life data to be reported as the data mature.
Regarding the use of chemotherapy in high-risk populations with early-stage endometrial cancer, the European Organisation for Research and Treatment of Cancer, along with the Nordic Society of Gynecologic Oncology, presented the results of a collaborative trial comparing external radiotherapy versus chemoradiotherapy in patients with stages I or II or IIIC1 (positive pelvic nodes only) and one of the following characteristics: grade 3 or myometrial invasion greater than 50%, DNA aneuploidy, or clear cell or serous histologic type.
Performance of a systematic lymphadenectomy was optional. A total of 375 patients were enrolled over a 10-year period and the study was closed early because of slow recruitment. The hazard ratio for progression-free survival was 0.58 (95% CI, 0.34%–0.99%) in favor of chemoradiation after a median followup of 3.5 years (221).
The Japanese GOG randomized patients to pelvic radiotherapy versus platinum-based chemotherapy in patients with stage IB to IIIC endometrial cancer and myometrial invasion greater than 50%. Assessment of para-aortic nodal status was performed in only 29% of cases. The investigators found no differences between the two experimental arms in terms of progression-free survival and overall survival. A subgroup analysis of 120 patients with either (a) stage IB tumor and age greater than 70 years or grade 3 endometrioid adenocarcinoma; or (b) stage II or IIIA tumor (positive cytologic findings) showed that chemotherapy was associated with significantly improved disease-free and overall survival rates (222).
The benefits achieved with combined chemoradiation, identification of the best chemotherapeutic regimen, and the identification of subgroups of patients who may benefit from these treatments deserves further investigation.
Clinical Stage II
Endometrial cancer involving the cervix either contiguously or by lymphatic spread has a poorer prognosis than disease confined to the corpus (192,223–226). Preoperative assessment of cervical involvement is difficult. Endocervical curettage has relatively high false-positive (50% to 80%) and false-negative rates. Histologic proof of cancer infiltration of the cervix or presence of obvious tumor on the cervix is the only reliable means of diagnosing cervical involvement, although ultrasonography, hysteroscopy, or MRI may show cervical invasion.
The relatively small number of true stage II cases in reported series and the lack of randomized, prospective studies preclude formulation of a definitive treatment plan. Three areas must be addressed in any treatment plan:
1. For optimal results, the uterus should be removed in all patients.
2. Because the incidence of pelvic lymph node metastases is about 36% in stage II endometrial cancer, any treatment protocol should include treatment of these lymph nodes.
3. Because the incidence of disease spread outside the pelvis to the paraaortic lymph nodes, adnexal structures, and upper abdomen is higher than in stage I disease, attention should be directed to evaluating and treating extrapelvic disease.
Two approaches usually were used in the treatment of clinical stage II disease:
1. Radical hysterectomy, bilateral salpingo-oophorectomy, and pelvic and para-aortic lymphadenectomy
2. Combined radiation and surgery (external pelvic radiation and intracavitary radium or cesium followed in 6 weeks by total abdominal hysterectomy and bilateral salpingo-oophorectomy)
An initial radical surgical approach to treatment of clinical stage II endometrial cancer has the advantage of collecting accurate surgicalpathologic information. Conversely, many patients with endometrial cancer are elderly and obese and have medical problems that make this approach unsuitable.
Reported results are no better than those with combined radiation and less radical surgical therapy (192). The use of radical hysterectomy may be limited to patients with anatomic problems that prevent optimum dosimetry or other conditions that conflict with the use of radiation therapy.
The most common, traditional approach to the management of clinical stage II endometrial cancer is to use external and intracavitary radiation followed by extrafascial hysterectomy. This combined approach resulted in 5- year survival rates of 60% to 80%, with severe gastrointestinal or urologic complications occurring in about 10% of patients (224,227,228). Patients who have medically inoperable disease are usually treated with external-beam radiation and one or two intracavitary insertions. Compared with combined radiation and surgery, the results with radiation alone are diminished, but about 50% of patients are long-term survivors (Table 37-9) (195).
Another method of management of clinical stage II endometrial cancer that is gaining favor is an initial surgical approach followed by radiation. This method is based on the difficulty in establishing the preoperative diagnosis of cervical involvement in the absence of a gross cervical tumor, the evidence that radiation is equally effective when given after hysterectomy, and the high incidence of extrapelvic disease when the cervix is involved. An extrafascial hysterectomy, bilateral salpingo-oophorectomy, peritoneal washings for cytology, and resection of grossly enlarged lymph nodes are performed. Modified radical or radical hysterectomy can be considered to obtain free margins. These procedures are followed by appropriate pelvic or extended-field external and intravaginal radiation, depending on the results of surgical staging. Excellent results were reported using this treatment scheme (225,229,230).
Clinical Stages III and IV
Clinical stage III disease accounts for about 7% to 10% of all endometrial carcinomas (231–233). Patients usually have clinical evidence of disease spread to the parametria, pelvic sidewall, or adnexal structures; less frequently, there is spread to the vagina or pelvic peritoneum. Treatment for stage III endometrial carcinoma must be individualized, but initial operative evaluation and treatment should be considered because of the high risk for occult lymph node metastases and intraperitoneal spread when disease is known to extend outside of the uterus into the pelvis. In the presence of an adnexal mass, the initial impetus for surgery is to determine the nature of the mass. Surgery is performed to determine the extent of disease and to remove the bulk of the disease if possible. This procedure should include peritoneal washings for cytologic examination, removal of bulky para-aortic and pelvic lymph nodes, biopsy or excision of any suspicious areas within the peritoneal cavity, and omentectomy and peritoneal biopsies. Except in patients with bulky parametrial disease, total abdominal hysterectomy and bilateral salpingo-oophorectomy should be performed. The goal of surgery is eradication of all macroscopic disease because this finding is of major prognostic importance in the management of patients with clinical stage III disease. Postoperative therapy can be tailored to the extent of disease.
Results of therapy depend on the extent and nature of disease. A 5-year survival rate of 54% was reported for all patients with stage III disease; however, the survival was 80% when only adnexal metastases were present, compared with 15% when other extrauterine pelvic structures were involved (231). Patients with surgical-pathologic stage III disease have a much better survival rate (40%) than those with clinical stage III disease (16%) (234). Patients who are treated with combined surgery and radiation fare better than patients who receive radiation therapy alone (233).
Stage IV endometrial adenocarcinoma, in which tumor invades the bladder or rectum or extends outside the pelvis, makes up about 3% of cases (233–236). Treatment of stage IV disease is patient dependent but usually involves a combination of surgery, radiation therapy, and systemic hormonal therapy or chemotherapy. One objective of surgery and radiation therapy is to achieve local disease control in the pelvis to provide palliative relief of bleeding, discharge, and complications involving the bladder and rectum. In one report, control of pelvic disease was achieved in 28% of 72 patients with stage IV disease treated with radiation alone or in combination with surgery, progestins, or both (235). Several reports noted a positive impact of cytoreductive surgery on survival, the median survival being about three times greater with optimal cytoreduction (18 to 34 months vs. 8 to 11 months, respectively) (235,236). Pelvic exenteration may be considered in the very rare patient in whom disease is limited to the bladder, rectum, or both (237,238).
Chemotherapy
In cases of advanced disease, chemotherapy with carboplatin and paclitaxel is now standard treatment (254–258). GOG209 demonstrated carboplatin and paclitaxel to be not inferior to the combination of paclitaxel, doxorubicin, and cisplatin, and the toxicity and tolerability profile favored carboplatin and paclitaxel (256). In a retrospective study, the overall response rate to combination carboplatin and paclitaxel in patients with either advanced or recurrent endometrial cancer was 43%, but only 5% achieved a complete response and the median survival was 13.2 months (255).
Single-agent chemotherapy options include paclitaxel, albumin-bound paclitaxel, docetaxel, carboplatin, cisplatin, doxorubicin, liposomal doxorubicin, and topotecan. When single-agents are used as first-line therapy, response rates range from 21% to 36%.
Recurrent Disease
About 15% of patients treated for early (stage I and II) endometrial cancer develop recurrent disease (239,240). In contrast, recurrent disease is detected in up to 50% of patients with advanced (stage III and IV) endometrial cancer (241). More than 50% of recurrences develop within 2 years and about 75% occur within 3 years of initial treatment. The distribution of recurrences is dependent in large part on the type of primary therapy, that is, surgery alone versus surgery plus local or regional radiotherapy. Vaginal and/or pelvic recurrences comprise over 50% of all recurrences in patients treated with surgery alone. Approximately 70% of recurrences in patients treated with combined surgery and radiotherapy (vaginal or external beam) are extrapelvic, most commonly lung, abdomen, lymph nodes (aortic, supraclavicular, inguinal), liver, brain, and bone (242,243).
At the time of diagnosis of recurrent endometrial cancer, approximately onethird of patients are asymptomatic and are only discovered to have recurrent cancer on routine surveillance examinations or radiologic studies. Almost onehalf of patients found to have lung metastases on follow-up chest imaging are asymptomatic. Approximately 90% of patients with metastases at other sites have symptoms related to the site of recurrence (244).
Patients with isolated vaginal recurrences fare better (70% 3-year survival) than those with pelvic recurrences (8% 3-year survival), or distant metastases (14% 3-
year survival) (240). Patients who initially have well-differentiated tumors or who
develop recurrent cancer more than 3 years after the primary therapy tend to have
an improved prognosis (245,246).
Surgery
A small subset of patients with isolated recurrent endometrial cancer may benefit from surgical intervention (247). A search for distant recurrences is obligatory, as such patients with metastatic disease are best treated with chemotherapy, unless there is an isolated pulmonary metastasis which may be amenable to excision. Recurrent disease isolated to the central pelvis following radiation is rarely seen. Selected patients with such a recurrence may be candidates for pelvic exenteration (248,249) although morbidity is high and survival rates are low. There is some evidence that radical pelvic resection in conjunction with intraoperative radiation may cure some patients, albeit with a high complication rate (250). Cytoreductive surgery for abdominal pelvic recurrences may be a treatment option for patients where complete gross surgical resection seems possible (251) (Fig. 37-6).
FIGURE 37-6 Radical pelvic sidewall resection. Photo taken of a patient with a left pelvic sidewall recurrence following resection of psoas muscle and bony ilium. Structures from lower to upper aspect of the photo (medial to lateral) include ureter, internal and external iliac artery, common iliac vein, lumbosacral trunk, obturator nerve (sacrificed), and femoral nerve. The cut edge of the psoas muscle is seen at left. (Provided by Sean C. Dowdy, MD, Mayo Clinic.)
Table 37-12 Carcinoma of the Endometrium: Stage Distribution and Actuarial Survival by Stage (Surgical and Clinical)
Patients with recurrent endometrial cancer are frequently elderly, overweight, and have numerous comorbidities thereby increasing operative and postoperative morbidity and mortality. Patients who are most often helped by surgical approaches to recurrent disease are younger, have well-differentiated endometrial tumors, have a longer time from original treatment to recurrence, and have negative margins or no gross residual disease after surgery (246,247).
Radiation Therapy
[8] Radiotherapy is the best treatment option for patients with isolated local– regional recurrences who have not received prior radiation (239,252,253). The best local control and subsequent cure are achieved by a combination of external pelvic radiation therapy followed by vaginal brachytherapy to deliver a total tumor dose of at least 6,000 cGy. Patients with low-volume disease limited to the pelvis (most of which is contained in the vagina) have the best outcomes.
Retrospective studies have shown complete remission rates after salvage radiotherapy for isolated vaginal relapse to be 40% to 80% in previously nonirradiated patients compared to 10% to 25% in patients who have received prior irradiation. Patients who undergo radiation for pelvic extension of disease have much lower reported survival rates of 0% to 26%. Factors associated with improved survival and control of pelvic disease in patients with locally recurrent endometrial cancer who receive radiotherapy include initial low-grade endometrioid histology, younger age at recurrence, recurrent tumor size 2 cm or less, time from initial treatment to recurrence of more than 1 year, vaginal versus pelvic disease, and use of vaginal brachytherapy.
Chemotherapy
The standard chemotherapy for recurrent metastatic endometrial cancer is the same as that for advanced stage disease, that is, combination chemotherapy with carboplatin and paclitaxel (257,258). However, there are no randomized controlled trials comparing carboplatin and paclitaxel in the recurrent setting. When single agents are used as second-line treatment, response rates range from 4% to 27%; paclitaxel being the most active in this setting. Temsirolimus and bevacizumab may be appropriate single-agent biologic therapy for patients who have progressed on previous cytotoxic therapy, although response rates are low at 4% and 13.5%, respectively (259). Studies have indicated that patients with programmed death ligand 1 (PD-L1) positive tumors and those with MMRdeficient (dMMR) or microsatellite instability high (MSI-H) endometrial cancer may respond to programmed death receptor-1 (PD-1) blockade with pembrolizumab (260,261). (See Tables 37-12 and 37-13.)
Table 37-13 Surgically Staged Endometrial Cancer: Actuarial 5-Year Survival Rate (%) by Histologic Grade and Stage (1988 Staging Criteria)
Grade
Stage 1 2 3
IA 93 90 69
IB 90 93 84
IC 89 81 63
IIA 91 78 57
IIB 78 75 58
IIIA 79 69 44
IIIB 77 40 21
IIIC 61 61 44
IVA — — 19
IVB 35 27 7
2415Adapted from Creasman WT, Odicino F, Maisonneuve P, et al. Carcinoma of the corpus
uteri. FIGO Annual Report on the results of treatment in gynecological cancer. J Epidemiol
Biostat 2001;6:45–86, with permission.
Table 37-14 Response to Progestin Therapy in Advanced or Recurrent Endometrial
Cancer
Hormonal Therapy
Hormonal therapy is an excellent management option for patients with asymptomatic recurrent endometrial cancer, especially for those who have lowgrade, hormone receptor positive tumors (244,246). Progestin therapy, either megestrol acetate 80 mg twice daily or medroxyprogesterone acetate 50 to 100 mg three times daily, is recommended when radiation therapy, surgery, or both, are not feasible for treatment of localized recurrent cancer or in nonlocalized disease. Progestins have reported response rates ranging from 15% to 25%.
Higher response rates are observed with well-differentiated tumors, expression of ER/PR receptors, longer disease-free interval, and location and extent of extrapelvic, especially pulmonary, metastatic disease. Progestin therapy should be given for at least 2 to 3 months before assessing response, and continued for as long as disease is static or in remission. In the presence of a relative contraindication to high-dose progestin therapy (e.g., prior or current thromboembolic disease, severe heart disease, or inability to tolerate progestin therapy), tamoxifen 20 mg twice daily is recommended, yielding response rates of about 10% to 20%. An alternating regimen of progestin and tamoxifen or an aromatase inhibitor may be substituted for a progestational agent, but response rates may be lower and toxicity higher. Failure to respond to hormonal therapy is an indication for initiation of chemotherapy. (See Table 37-14.)
UTERINE SARCOMA
Uterine sarcomas are relatively rare tumors arising from mesenchymal tissues of the uterus, including the uterine muscle and endometrial stroma. They account for approximately 3% of all uterine malignancies (262). In general, [9] uterine sarcomas are some of the most malignant uterine tumors and differ from endometrial carcinomas with regard to diagnosis, clinical behavior, pattern of spread, and management (265).
Classification and Staging
The two most common histologic types of uterine sarcoma are leiomyosarcoma (LMS) and endometrial stromal sarcoma (ESS), which make up approximately 60% and 20% of uterine sarcomas, respectively (Table 37-15). Rarer subtypes of malignant mesenchymal tumors that occur in the uterus are undifferentiated uterine sarcoma (UUS), adenosarcoma, and perivascular epithelioid cell neoplasm (PEComa) (264–266). Carcinosarcomas, also known as malignant mixed müllerian tumors (MMMT), were previously categorized and included in uterine sarcoma treatment algorithms, but are thought to represent high-grade carcinomas with mesenchymal metaplasia (267).
Staging of uterine sarcomas had previously been based on the FIGO staging system for endometrial adenocarcinomas. Recognition of the unique spread pattern and clinical behavior of uterine sarcomas required an independent staging system for each histologic type that was developed for LMS and ESS in 2009
(Tables 37-16 and 37-17) (268).
Endometrial Stromal Sarcoma
ESSs occur primarily in perimenopausal women between ages 45 and 50 years, with approximately one-third occurring in postmenopausal women. There is no apparent relationship to parity or associated diseases, and these tumors are rare in African American women. The most frequent symptom is abnormal uterine bleeding; abdominal pain, and pressure caused by an enlarging uterus occur less often, and some patients do not have any symptoms. Pelvic examination usually reveals regular or irregular uterine enlargement, sometimes associated with rubbery parametrial induration. The diagnosis may be determined by endometrial biopsy, but the usual preoperative diagnosis is uterine leiomyoma (267).
ESSs are composed of cells resembling normal endometrial stroma in the proliferative phase. They are divided into two types based on histologic criteria, such as cellular atypia, mitotic activity, and vascular invasion: low-grade ESS and high-grade ESS including undifferentiated sarcoma. Molecular profiling supports dividing these endometrial sarcomas into low- and high-grade entities based on histopathology and clinical behavior and outcomes (266).
Low-grade ESS is distinguished microscopically by mitotic rates of less than 10 MF/10 HPF, slight nuclear atypia, minimal necrosis, and presence of estrogen and progesterone receptors; myometrial and lymph–vascular invasion are common. Up to one-half of these tumors harbor a JAZF1–SUZ12 gene fusion, which correlates with lower-grade, earlier-stage tumors and may be helpful in diagnosing these tumors and differentiating them from higher-grade endometrial stroma sarcomas (266,269).
Table 37-15 Classifications of Uterine Sarcomas
I. Pure nonepithelial tumors
A. Homologous
1. Endometrial stromal tumors
a. Low-grade stromal sarcoma
b. High-grade or undifferentiated stromal sarcoma
2. Smooth muscle tumors
a. Leiomyosarcoma
b. Leiomyoma variants
1. Cellular leiomyoma
2. Leiomyoblastoma (epithelioid leiomyoma)
c. Benign metastasizing tumors
1. Intravenous leiomyomatosis
2. Benign metastasizing leiomyoma
3. Disseminated peritoneal leiomyomatosis
B. Heterologous
1. Rhabdomyosarcoma
2. Chondrosarcoma
3. Osteosarcoma
24184. Liposarcoma
II. Mixed epithelial–nonepithelial tumors
A. Malignant mixed müllerian tumor
1. Homologous (carcinosarcoma)
2. Heterologous
B. Adenosarcoma
Modified from Clement P, Scully RE. Pathology of uterine sarcomas. In: Coppleson M, ed. Gynecologic oncology: Principles and Practice. New York: Churchill-Livingston, 1981, with permission.
Low-grade ESSs are found to extend beyond the uterus in 40% of cases at the time of diagnosis, but the extrauterine spread is confined to the pelvis in twothirds of these cases. Upper abdominal, pulmonary, and lymph node metastases are less common. Recurrences occur in almost one-half of cases at an average interval of 5 years after initial therapy and local recurrence is more common than distant metastasis. These tumors tend to have protracted clinical courses (263– 267,269,270).
Optimal initial therapy for patients with low-grade ESS consists of surgical excision of all grossly detectable tumors. Total hysterectomy should be performed. The fallopian tubes and ovaries should be removed because of the propensity for tumor extension into the parametria, broad ligaments, adnexal structures, and the possible stimulating effect of estrogen on tumor cells if the ovaries are retained. The relatively low rate of lymph node metastasis in the absence of gross lymph node involvement or extrauterine disease suggests that lymph node dissection is not necessary. Pelvic irradiation is recommended for inadequately excised or locally recurrent pelvic disease. Owing to the indolent nature of the disease, cytotoxic chemotherapy is unlikely to be beneficial.
Hormonal therapy with an aromatase inhibitor or progestin yields response rates close to 50% in patients with recurrent or metastatic disease. Recurrent or metastatic lesions may be amenable to surgical excision. Prolonged survival and cure are common even after the development of recurrent or metastatic disease with the overall 5-year survival exceeding 90% (263–267,270).
High-grade ESS and UUS are highly malignant neoplasms, although the prognosis is poorer for the latter. Histologically, both exhibit greater than 10 MF/10 HPF, necrosis, significant cytologic pleomorphim, and prominent myometrial and lymph–vascular invasion. UUS, by definition, lacks recognizable morphologic or immunohistochemical evidence of smooth muscle or endometrial stromal differentiation. Higher-grade ESSs have been found to have a unique genetic rearrangement YWHAE–FAM22A/B which correlates with a more aggressive clinical behavior (266,269).
Table 37-16 FIGO Staging for Uterine Endometrial Stromal Sarcoma and
Adenosarcoma
Stage Definition
I Tumor limited to uterus IA: tumor limited to
endometrium/endocervix with no
myometrial invasion
IB: ≤½ myometrial invasion
IC: >½ myometrial invasion
II Tumor extends beyond the uterus,
within the pelvis
IIA: adnexal involvement
IIB: involvement of other pelvic tissues
III Tumor invades abdominal tissues (not
just protruding into the abdomen)
IIIA: one site
IIIB: more than one site
IIIC: metastasis to pelvic and/or paraaortic lymph nodes
IV IVA: tumor invades bladder and/or
rectum
IVB: distant metastasis
Prat J. FIGO staging for uterine sarcomas. Int J Gynecol Obstet 2009;104:177–178.
Table 37-17 FIGO Staging for Uterine Leiomyosarcoma
Stage Definition
I Tumor limited to uterus IA: <5 cm
IB: >5 cm
II Tumor extends beyond the uterus, within
the pelvis
IIA: adnexal involvement
IIB: involvement of other pelvic
tissues
III Tumor invades abdominal tissues (not just
protruding into the abdomen)
IIIA: one site
IIIB: more than one site
IIIC: metastasis to pelvic and/or
para-aortic lymph nodes
2420IV IVA: tumor invades bladder
and/or rectum
IVB: distant metastasis
Prat J. FIGO staging for uterine sarcomas. Int J Gynecol Obstet 2009;104:177–178.
Treatment of high-grade ESS and UUS should consist of total hysterectomy and bilateral salpingo-oophorectomy. The poor therapeutic results with surgery alone, suggest that adjuvant radiation therapy, chemotherapy, or both, should be used in combination with surgery, however, there is no data to support this. These tumors are not responsive to hormonal therapy. The 5-year disease-free survival rate for these tumors is 25% (263–267,270).
Uterine tumor resembling ovarian sex-cord tumor (UTROSCT) is a rare variant of ESS, which contains elements resembling ovarian sex-cord tumors. Immunohistochemically, these tumors express calretinin, inhibin, CAM5.2, and progesterone receptor; positive staining for calretinin and at least one of the other markers confirms the diagnosis of UTROSCT. The mean age of diagnosis is 52.
The most common presenting symptom is abnormal uterine bleeding. Although some of these tumors are large and have infiltrative margins, almost all behave benignly. Endometrial stromal tumors with sex-cord–like elements (ESTSCLE) tend to behave more aggressively with larger tumors, more extrauterine metastases, and a higher rate of recurrence (271).
Leiomyosarcoma
The median age for women with LMS is 54 (range 43 to 63) years, and premenopausal women have a better survival. This malignancy has no relationship with parity, and the incidence of associated diseases is not as high as in endometrial carcinomas. African Americans have a higher incidence and a poorer prognosis than women of other races. A history of prior pelvic radiation can be elicited in about 4% of women with uterine LMS. The incidence of sarcomatous change in benign uterine leiomyomas is reported to be between 0.13% and 0.81%, and most uterine LMSs are unassociated with pre-existing leiomyomas (272–276).
Presenting symptoms, which are usually of short duration (mean = 6 months) and not specific to the disease, include vaginal bleeding, pelvic pain or pressure, and awareness of an abdominal mass. The principal physical finding is the presence of a pelvic mass. The diagnosis should be suspected if pelvic pain accompanies an enlarging uterus, especially in a postmenopausal woman.
Endometrial biopsy, although not as useful as in ESSs or carcinomas, may establish the diagnosis in as many as one-third of cases if the lesion is submucosal. MRI, especially contrast-enhanced MRI, may be more diagnostic than ultrasound or CT in differentiating LMSs and smooth muscle tumors of uncertain malignant potential (STUMP) from leiomyomas (275).
Pathologically, the number of mitoses in uterine smooth muscle tumors has traditionally been considered the most reliable indicator of malignant behavior. Tumors with less than 5 MF/10 HPF behave in a benign fashion and tumors with more than 10 MF/10 HPF are frankly malignant, while tumors with 5 to 10 MF/10 HPF-termed STUMP have a less predictable behavior. In addition to mitotic index greater than 10, other histologic indicators used to classify uterine smooth muscle tumors as malignant are severe cytologic atypia, infiltrating borders, coagulative tumor cell necrosis, and presence of vascular invasion (266,277) (Fig. 37-7). Gross presentation of the tumor at the time of surgery is an important prognostic indicator. Tumors greater than 5 cm, with infiltrating rather than pushing borders, and extension beyond the uterus have a poorer prognosis.
The pattern of tumor spread is usually through the myometrium, into pelvic blood vessels and lymphatics, contiguously to pelvic structures and intra-abdominally, and distantly (most often the lungs). Reported survival rates for women with uterine LMS range from 20% to 63% (mean 47%). Younger women with stage I disease do best, while older women and those with advanced disease do least well (272–276).
Surgery is the mainstay of treatment for uterine LMS. For women with uterineconfined disease, total hysterectomy is recommended, optimally avoiding intraoperative morcellation which may increase recurrence and death in women with previously undiagnosed uterine LMS (275,278). Bilateral salpingo-oophorectomy should be done in postmenopausal women and in women with gross extrauterine disease. Preservation of the ovaries in premenopausal women does not increase the risk of recurrence. Retroperitoneal lymphatic spread is rare in women with early-stage disease and lymphadenectomy is not associated with a survival advantage. In women with disease spread outside the uterus, complete cytoreductive surgery or metastasectomy (especially in the lungs) may improve survival (272–276).
FIGURE 37-7 Leiomyosarcoma of the uterus. Interlacing bundles of spindle cells have fibrillar cytoplasm, irregular and hyperchromatic nuclei, and multiple mitotic figures. (Provided by Gordana Stevanovic, MD, and Jianyu Rao, MD, Department of Pathology, UCLA.)
Radiation therapy does not improve outcomes for women with uterine LMS. Data from a phase III trial conducted by the European Organization for Research and Treatment of Cancer (EORTC) comparing adjuvant postoperative pelvic radiation therapy to observation only for women with early-stage uterine LMS found no differences in overall survival or recurrence rates (279). This lack of benefit of adjuvant postoperative radiotherapy was supported by results from a Surveillance, Epidemiology, and End Results (SEER) database (280). Therefore, postoperative adjuvant radiation therapy is not recommended for completelyresected, early-stage disease. Chemotherapy is indicated and increases survival in patients with advanced LMS. Several single agents, such as dacarbazine, doxorubicin, epirubicin, eribulin, gemcitabine, ifosfamide, liposomal doxorubicin, pazopanib, temozolomide, and trabectedin have been shown to have some activity in treating advanced or metastatic LMSs yielding response rates of 15% to 25%, and doxorubicin is the most active. Combination chemotherapy with gemcitabine/docetaxel or doxorubicin/ ifosfamide +/– dacarbazine has yielded response rates of 53% and 30%, respectively, in advanced uterine LMS, however, median time to progression is usually less than 6 months and survival less than 2 years (259). A phase II trial of the combination of doxorubicin/olaratumab versus single-agent doxorubicin for the treatment of soft tissue sarcomas including LMSs found the combination to be superior with a median progression-free survival of 6.6 versus 4.1 months and a median overall survival of 26.5 versus 14.7 months, respectively (281). Adjuvant postoperative chemotherapy for early-stage uterine LMS has failed to improve survival or lower the 40% to 70% risk of recurrence, despite its increasing use (282).
Five other clinical pathologic variants of uterine smooth muscle tumors deserve special comment: myxoid LMS, epithelioid LMS, intravenous leiomyomatosis, benign metastasizing uterine leiomyoma, and disseminated peritoneal leiomyomatosis (266,277).
Myxoid LMS is characterized grossly by a gelatinous appearance and apparent circumscribed border. Microscopically, the tumors have a myxomatons stroma and extensively invade adjacent tissue and blood vessels. The mitotic rate may be low (0 to 2 MF/10 HPF), which belies their aggressive behavior and poor prognosis (283). Surgical excision by hysterectomy is the mainstay of treatment.
The low mitotic rate and abundance of intracellular myxomatous tissue suggest that these tumors would not be responsive to radiation therapy or chemotherapy.
Epithelioid LMS often forms large fleshy intrauterine masses, commonly associated with hemorrhage. They exhibit a microscopic pattern of epithelioid differentiation consisting of round cells with eosinophilic and granular or clear cytoplasm growing in sheets, nests, or cords in addition to cytologic atypia, increased mitotic activity (>5MF/10 HPF), and tumor cell necrosis. These tumors must be distinguished from placental site trophoblastic tumors and epithelioid trophoblastic tumors. Epithelioid LMSs are aggressive tumors with a prognosis similar to myxoid LMSs, with a propensity for late recurrences. Hysterectomy is the only treatment proven to be of any benefit (277).
Intravenous leiomyomatosis is characterized by the growth of histologically benign smooth muscle into venous channels within the broad ligament and then into uterine and iliac veins. The intravascular growth takes the form of visible, worm-like projections that extend from a myomatous uterus into the parametria toward the pelvic sidewalls. It may be confused with low-grade stromal sarcoma.
Symptoms are related to the associated uterine myomas. Most women are in the late fifth and early sixth decades of life. The prognosis is excellent, even when tumor is left in pelvic vessels. Late local recurrences can occur, and deaths from extension into the inferior vena cava or metastases to the heart have been reported. Estrogen may stimulate the proliferation of these intravascular tumors.
Treatment should be total hysterectomy with bilateral salpingo-oophorectomy and excision of any extrauterine tumor when technically feasible. Antiestrogen therapy is appropriate if residual tumor remains after surgery (277). Benign metastasizing uterine leiomyoma is a rare condition in which a histologically benign uterine smooth muscle tumor acts in a somewhat malignant fashion and produces benign metastasis, usually to the lungs or lymph nodes. In most instances, intravenous leiomyomatosis is not apparent. The metastasizing myomas are capable of growth at distant sites, whereas the intravenous tumors spread only by direct extension within blood vessels. Experimental and clinical evidence suggests that these tumors are stimulated by estrogen. Removing the source of estrogen, by oophorectomy or withdrawal of exogenous estrogen, or by treatment with an antiestrogen or a gonadotropin agonist has an ameliorating effect. Surgical treatment should consist of total hysterectomy, bilateral salpingooophorectomy, and resection of pulmonary metastases, if possible (284).
Disseminated peritoneal leiomyomatosis is a rare clinical entity characterized by benign smooth muscle nodules scattered throughout the peritoneal cavity. This condition probably arises as a result of metaplasia of subperitoneal mesenchymal stem cells to smooth muscle, fibroblasts, myofibroblasts, and decidual cells under the influence of estrogen and progesterone. Most reported cases occurred in 30- to 40-year-old women who are or were recently pregnant or who have a long history of oral contraceptive use. Intriguing features of this disease are its grossly malignant appearance, benign histology, and favorable clinical outcome.
Intraoperative diagnosis requires frozen-section examination. Extirpative surgery, including hysterectomy, bilateral salpingo-oophorectomy, omentectomy, and excision of as much gross-tumor as possible, may be indicated in menopausal women. Removal of the source of excess estrogen and treatment with progestins or antiestrogens has resulted in regression of unresected tumor masses. Almost all women have a good prognosis (285).
Adenosarcoma
Adenosarcoma consists of an admixture of benign-appearing neoplastic glands and a sarcomatous stroma. It accounts for only about 5% of all uterine sarcomas. Most patients present with postmenopausal uterine bleeding, and the disease is diagnosed or suspected on endometrial sampling. Most adenosarcomas are well circumscribed and are limited to the endometrium or superficial myometrium. Standard treatment is total hysterectomy and bilateral salpingo-oophorectomy; routine lymph node dissection is not indicated. Because recurrences, mostly in the form of local pelvic or vaginal disease, are reported to occur in up to 40% to 50% of cases, adjuvant postoperative vaginal brachytherapy or pelvic radiation has been recommended. There is little evidence for the use of adjuvant chemotherapy or hormonal therapy. Survival is mainly impacted by the presence or absence of sarcomatous overgrowth, which is found in about 30% of cases. Overall 5-year survival is reported to be 70% to 80% without sarcomatous overgrowth versus 50% to 60% with sarcomatous overgrowth (286).
Nhận xét
Đăng nhận xét