Carcinoma of the epithelial lining (endometrium) of the uterine corpus is the most common female pelvic malignancy. The American Cancer Society estimates that 37,400 cases of this cancer will occur in the year 2000 in the United States. Factors influencing its prominence are the declining incidence of cervical cancer, longer life expectancy, and earlier diagnosis.
Adenocarcinoma of the endometrium, the most prevalent histologic subtype, is currently the fourth most common cancer in women, ranking behind breast, bowel, and lung cancers. Endometrial adenocarcinoma is the seventh leading cause of death from malignancy in females, accounting for 6,300 deaths each year.
Age Endometrial cancer is primarily a disease of postmenopausal women, although 25% of cases occur in premenopausal patients, with 5% of cases developing in patients < 40 years old.
Geography The incidence of endometrial cancer is higher in western nations and very low in eastern countries.
Immigrant populations tend to assume the risks of native populations, highlighting the importance of environmental factors in the genesis of this disease. Endometrial cancers tend to be more common in urban than rural residents. In the United States, white women have a twofold higher incidence than black women.
Adenocarcinoma of the endometrium may arise in normal, atrophic, or hyperplastic endometrium. Two mechanisms are generally believed to be involved in the development of endometrial cancer. In approximately 75% of women, there is a history of exposure to unopposed estrogen, either endogenous or exogenous. The tumors in these women begin as endometrial hyperplasia and progress to carcinomas, which usually are better differentiated and have a favorable prognosis.
In the other 25% of women, carcinomas appear spontaneously and are not clearly related to a transition from atypical hyperplasia, but rather, arise in a background of atrophic or inert endometrium. These neoplasms tend to be associated with a more undifferentiated cell type and a poorer prognosis.
Unopposed estrogen It has been hypothesized that long-term estrogenic stimulation of the endometrium unmodified by progesterone has a role in the development of endometrial carcinoma. This hypothesis derives from observations that women who are infertile or obese or who have dysfunctional bleeding due to anovulation are at high risk for this disease, as are women with estrogen-secreting granulosa theca cell ovarian tumors. Also, the recognition that atypical adenomatous (complex) hyperplasia is a precursor of cancer, and that it is associated with unopposed estrogen in women, underscores the importance of the association among risk factors, estrogens, and cancer. In the late 1970s and early 80s, several case-control studies demonstrated that the risk of endometrial cancer is increased 4- to 15-fold in long-term estrogen users, as compared with age-matched controls.
Diet The high rate of occurrence of this disease in western societies and the very low rate in eastern countries suggest a possible etiologic role for nutrition, especially the high content of animal fat in western diets. There may be a relationship between high-fat diets and the higher incidence of endometrial carcinoma in women with conditions of unopposed estrogen: Endogenous estrogens rise in postmenopausal women because of increased production of androstenedione or a greater peripheral conversion of this hormone to estrone. In obese women, the extraglandular aromatization of androstenedione to estrone is increased in fatty tissue.
Obesity Phenotypically, the majority of women who develop endometrial cancer tend to be obese. Women who are 30 pounds over ideal weight have a 3-fold increased risk of developing endometrial cancer, while those 50 pounds or more over ideal weight have a 10-fold increased risk.
Parity Nulliparous women are at 2 times greater risk of developing endometrial cancer, females who undergo menopause after age 52 are at 2.5 times greater risk, and those who experience increased bleeding at the time of menopause are at 4 times greater risk.
Other risk factors Other known risk factors for endometrial cancer include diabetes mellitus, hypertension, endometrial hyperplasia, a family history of endometrial cancer, and use of exogenous hormones. Diabetic females have a 3-fold increased risk, and hypertensive patients, a 1.5-fold greater risk. Whereas patients found to have simple endometrial hyperplasia have a very low risk of progression to cancer, 29% of those with complex atypical hyperplasia, if left untreated, will develop adenocarcinoma.
Tamoxifen (Nolvadex), a nonsteroidal antiestrogen commonly used in the management of breast carcinoma, exerts its primary effect by blocking the binding of estrogen to estrogen receptors. It also exerts mild estrogenic effects on the female genital tract. This weak estrogenic effect presumably accounts for an increased frequency of endometrial carcinoma observed in women receiving prolonged adjuvant tamoxifen therapy for breast carcinoma.
Initially reported in 1985, the increased frequency of endometrial carcinoma in patients treated with tamoxifen was characterized more fully in a study of 1,846 women recorded in the Swedish Cancer Registry. This study reported a 6.4-fold increase in the relative risk of endometrial carcinoma with a daily dose of 40 mg. The greatest cumulative risk was observed after 5 years of tamoxifen use.
The National Surgical Adjuvant Breast and Bowel Project (NSABP) subsequently reported on the incidence of other cancers in 2,843 women with node-negative, estrogen-receptorpositive breast cancer treated with either tamoxifen or placebo in its B-14 randomized trial and an additional 1,220 patients treated with tamoxifen in another NSABP trial. The relative risk of endometrial carcinoma in the tamoxifen-treated patients was 7.5. The hazard rate was 0.2 cases per 1,000 with placebo and 1.6/1,000 with tamoxifen therapy. The mean duration of tamoxifen therapy for all patients was 35 months, and 36% of the cancers had developed by 2 years after the initiation of treatment.
These data raise the question as to whether tamoxifen should be used as adjuvant therapy in women at relatively low risk for breast cancer recurrence. First, it should be recognized that the endometrial cancers that develop in patients receiving tamoxifen exhibit the same stage, grade, and prognosis distribution as other endometrial cancers. This means that the cure rate should be high. Second, adjuvant tamoxifen reduces the cumulative rate of recurrence of breast cancer from 228 to 124 cases per 1,000 and the cumulative rate of second primary breast cancers from 40.5 to 23.5 cases per 1,000.
When all of these facts are taken into account, there is an overall 38% reduction in the cumulative hazard rate in tamoxifen-treated patients. Thus, the benefits of tamoxifen outweigh the risks of endometrial cancer.
If tamoxifen is to be used, how should treated women be screened? The current recommendations are to educate patients about the significance of abnormal spotting, bleeding, or discharge and to investigate promptly any of these abnormalities.
Some experts have proposed that tamoxifen-treated women be screened with transvaginal ultrasound. However, recent data suggest a high false-positive rate and low frequency of significant findings, leading to the conclusion that endometrial screening is not warranted.
Postmenopausal women Symptoms of early endometrial carcinoma are few. However, 90% of patients complain of abnormal vaginal discharge, and 80% of these women experience abnormal bleeding, usually after menopause. In the general population, 15% of postmenopausal women presenting with abnormal bleeding will be found to have endometrial carcinoma. Signs and symptoms of more advanced disease include pelvic pressure and other symptoms indicative of uterine enlargement or extrauterine tumor spread.
Premenopausal women The diagnosis of endometrial cancer may be more difficult to make in premenopausal patients. The physician must maintain a high index of suspicion in this group of patients and perform endometrial sampling in any women who complain of prolonged, heavy menstrual periods or intermenstrual spotting.
Screening There is no role for screening of asymptomatic patients for endometrial cancer.
Outpatient endometrial sampling procedures, such as endometrial biopsy or aspiration curettage coupled with endocervical sampling, are definitive if positive for cancer. The results of endometrial biopsies correlate well with endometrial curettings, and these biopsy procedures have the advantage of avoiding general anesthesia. However, if sampling techniques fail to provide sufficient diagnostic information or if abnormal bleeding persists, a formal dilation and curettage is required.
Dilation and curettage The gold standard for assessing uterine bleeding and diagnosing endometrial carcinoma is formal fractional dilation and curettage. Before dilating the cervix, the endocervix should be curetted. Next, careful sounding of the uterus is accomplished. Dilation of the cervix is then performed, followed by systematic curetting of the entire endometrial cavity. Cervical and endometrial specimens should be kept separate and forwarded for pathologic interpretation.
Adenocarcinoma Endometrioid adenocarcinoma is the most common form of endometrial carcinoma, comprising 75%-80% of cases. It varies from well differentiated to undifferentiated The former demonstrates well-preserved glands in at least 95% of the tumor, whereas, in the latter, less than half of the neoplasm shows glandular differentiation. Squamous differentiation can be seen in 30%-50% of cases.
Adenoacanthoma Adenocarcinoma with benign squamous differentiation has been termed adenoacanthoma and generally has a good prognosis.
Adenosquamous carcinoma If the squamous component resembles squamous carcinoma, the tumor is designated an adenosquamous carcinoma. These lesions tend to have a worse prognosis due to their association with a poorly differentiated glandular component.
Serous carcinoma is a very aggressive form of endometrial cancer that accounts for < 10% of these tumors. Serous cancer of the endometrium closely resembles serous carcinoma of the ovary and fallopian tube and is usually found in an advanced stage in older women.
Clear-cell carcinomas of the endometrium closely resemble their counterparts in the cervix, vagina, and ovary. As with serous cancers, these tumors generally occur in older women and have a poor prognosis due to their propensity for early intraperitoneal spread.
Secretory adenocarcinoma is an uncommon endometrial cancer that resembles secretory endometrium with its associated progestational changes. These cancers tend to be of low grade and have a good prognosis.
Two large prospective Gynecologic Oncology Group (GOG) surgical staging trials reported in 1984 and 1987 helped define the prognostic factors for endometrial carcinoma and the current treatment approach. In addition to evaluating the predictive value of such factors as age, race, and endocrine status, the studies confirmed that prognosis is directly related to the presence or absence of easily determined uterine and extrauterine risk factors.
Uterine prognostic factors include histologic cell type, tumor grade, depth of myometrial invasion, occult extension to the cervix, and vascular space invasion.
Extrauterine prognostic factors include adnexal metastases, intraperitoneal spread to other extrauterine structures, positive peritoneal cytology, pelvic lymph node metastases, and aortic node involvement.
Uterine size was previously believed to be a risk factor and was part of the older clinical staging system. However, recent information indicates that uterine size is not an independent risk factor, but rather, relates to cell type, grade, and myometrial invasion.
Surgical staging Cell type and grade can be determined before hysterectomy, although in some series, grade, as determined by dilation and curettage, has an overall 31% inaccuracy rate compared with grade in the hysterectomy specimen, and grade 3 tumors have a 50% inaccuracy rate. Recognition of all of the other factors requires an exploratory laparotomy, peritoneal fluid sampling, and hysterectomy with careful pathologic interpretation of all removed tissue. This primary surgical approach led the International Federation of Gynecology and Obstetrics (FIGO) in 1988 to define endometrial cancer as a surgically staged disease, incorporating many of the prognostic factors into the staging process.
Approximately 90% of patients with a diagnosis of endometrial cancer are medically able to undergo surgery. Preparation for this surgery should include evaluation of such concurrent medical problems as hypertension and diabetes, which are frequently found in patients with endometrial cancer.
Open surgical procedure The operative procedure is performed through an adequate abdominal incision that allows for thorough intra-abdominal exploration and retroperitoneal lymph node removal if necessary. On entry into the peritoneal cavity, fluid samples are obtained for subsequent cytologic determination (intraperitoneal cell washings). Next, thorough intra-abdominal and pelvic exploration is undertaken, with biopsy or excision of any suspicious lesions. In particular, the uterus should be observed for tumor breakthrough of the serosal surface. The distal ends of the fallopian tubes are clipped or ligated to prevent possible tumor spill during uterine manipulation.
These procedures should be followed by total extrafascial hysterectomy and bilateral salpingo-oophorectomy. The excised uterus is opened away from the operating table, and the depth of myometrial penetration and presence or absence of endocervical involvement are determined by clinical observation or microscopic frozen section. The depth of myometrial invasion can be accurately assessed in over 90% of cases.
Laparoscopic surgery An alternative method of surgically staging patients with clinical stage I endometrial cancer is gaining in popularity. This approach combines laparoscopically assisted vaginal hysterectomy with laparoscopic lymphadenectomy.
Childers and colleagues described their experience with laparoscopic surgery in 59 patients with clinical stage I endometrial carcinoma. The peritoneal cavity was thoroughly inspected, intraperitoneal washings were obtained, and a laparoscopically assisted vaginal hysterectomy was performed. Laparoscopic pelvic and aortic lymph nodes were sampled in all patients with grade 2 or 3 lesions, as well as those with grade 1 lesions who were found to have > 50% myometrial invasion on frozen-section analysis. In two patients, laparoscopic lymphadenectomy was precluded by obesity.
Six patients noted to have intraperitoneal disease at laparoscopy underwent exploratory laparotomy. Two additional patients required laparotomy for complications, including a transected ureter and a cystotomy. The mean hospital stay was 2.9 days.
Laparoscopic assisted surgical staging is feasible in a select group of patients. However, it is not yet known whether this approach is applicable to all patients with clinical stage I disease. In particular, patients who are overweight or have intra-abdominal adhesions may not be ideal candidates. Para-aortic lymphadenectomy is technically more difficult through the laparoscope. To obtain adequate exposure, it is necessary to elevate the small bowel mesentery into the upper abdomen, which becomes increasingly difficult as the patients weight increases, especially when weight exceeds 180 pounds.
Lymph node sampling Any suspicious pelvic or para-aortic lymph nodes should be removed for pathologic evaluation. If there is no gross residual intraperitoneal tumor, pelvic and para-aortic lymph nodes should be sampled for the following indications:
Lymph nodes need not be sampled in patients whose tumor is limited to the endometrium, regardless of grade, because < 1% of these patients have disease spread to pelvic or para-aortic lymph nodes. The decision of whether to perform lymph node sampling is less clear-cut for patients whose only risk factor is invasion of the inner half of the myometrium, particularly if tumor grade is 1 or 2. This group has a £ 5% chance of node positivity.
Sampling procedures When indicated, para-aortic node sampling can be performed through a midline peritoneal incision over the common iliac arteries and aorta. Alternatively, sampling can be performed on the right by mobilizing the right colon medially and on the left by mobilizing the left colon medially. In each case, a sample of lymphatics and lymph nodes is resected along the upper common iliac vessels on either side and from the lower portion of the aorta and vena cava. On the left side, the lymph nodes and lymphatics are slightly posterior to the aorta, and on the right side, they lie primarily in the vena caval fat bed.
In cases where pelvic lymph node sampling is indicated, a sample of lymph nodes is taken from the distal common iliac artery, the superior iliac artery and vein, and the group of nodes that lie along the obturator nerve. In a lymph node sampling procedure, it is important to try to achieve an adequate sample of nodes from each anatomic site, but no attempt is made to perform a complete lymphadenectomy.
Surgical staging After these procedures, the patient is surgically staged according to the 1988 FIGO criteria. The overall surgical complication rate after this type of staging is approximately 20%. The rate of serious complications is 6%, and includes vascular, ureteral, and bowel injury.
Adjuvant radiation therapy
Preoperative vs postoperative irradiation Historically, adjuvant radiation treatment has been administered preoperatively in the majority of women with a diagnosis of endometrial cancer. Now, however, many US medical facilities offer surgical staging and postoperative adjuvant irradiation to those women with poor prognostic factors. In retrospective reports, postoperative radiotherapy seems to compare favorably with preoperative irradiation. In one report, 5-year survival rates for patients with stage I, grade 3 tumors treated with preoperative or postoperative irradiation were 64% and 73%, respectively.
Postoperative irradiation seems preferable to preoperative irradiation because patients can be more properly selected, and techniques can be altered to suit the prognostic factors found at the time of surgery. In addition, patients without poor prognostic factors can be spared unnecessary irradiation.
Postoperative irradiation is known to decrease the incidence of locoregional recurrences. A cuff recurrence rate of 7%-14% is predictable for all stage I patients after surgery alone for endometrial carcinoma, although certain subgroups with more risk factors may have a higher incidence of recurrence. Several retrospective reports suggest a decrease in vaginal recurrences after radiotherapy.
Pelvic irradiation, vaginal irradiation, or both? Most adjuvant radiation has been delivered using external-beam irradiation directed to the pelvis, which allows for treatment of the pelvic nodes.
A randomized trial conducted by the GOG compared the results of pelvic radiation to observation following hysterectomy and lymphadenectomy. The 2-year relapse-free survival rate was significantly higher among patients treated with pelvic radiation than among those who were observed (96% vs 88%). Overall survival rates at 3 years did not differ significantly between the two groups. Criticism of this trial includes the fact that the majority of patients had low-grade histology, as well as shallow myometrial penetration. The predominant site of failure in these patients was the vagina (13 of 17 recurrences, with the remaining 4 in the pelvis).
A few reports have demonstrated excellent local control with vaginal radiation alone. Eltabbakh et al documented excellent local control in 303 patients with grade 1 or 2 tumors and < 50% myometrial penetration who were treated with intracavitary radiation to the vagina only.
In a randomized study from the Norwegian Radium Hospital, pelvic irradiation significantly decreased the incidence of locoregional recurrences compared with vaginal irradiation alone. Patients with deeply invasive, grade 3 tumors had lower death and recurrence rates when treated with pelvic plus vaginal irradiation, as compared with vaginal irradiation alone.
Many women with endometrial cancer are being treated with lymphadenectomy at the time of hysterectomy. There has been some interest in using vaginal radiation alone to treat women who have negative nodes but deep myometrial penetration or high-grade histology. Two small reports have demonstrated excellent outcomes for such patients, However, more experience is needed to determine whether vaginal radiation alone is adequate.
In general, vaginal irradiation alone is reserved for patients at low risk for pelvic node metastasis. Because of increased rectal and vaginal sequelae, treatment of the entire length of the vagina is usually not recommended.
Pelvic and vaginal radiation have been combined for the adjuvant treatment of some patients. Patients with cervical involvement or extrauterine disease, who may have an increased incidence of local failure, may benefit from the two treatments combined, although there are no data to suggest that the addition of brachytherapy improves outcome over external-beam radiation alone. Patients with uterine-confined disease have excellent local control following treatment with either type of radiation. Combining the two treatments has not been shown to benefit these patients.
HDR vs LDR vaginal irradiation Vaginal radiation can be delivered with high-dose rate (HDR) or low-dose rate (LDR) equipment. Both techniques have resulted in excellent local control rates and low morbidity when administered by experienced practitioners. Each technique has its advantages. HDR treatments require multiple insertions, generally with one insertion done every week for 3-6 weeks. However, hospitalization is not required, and each insertion takes only a brief amount of time. LDR treatments are delivered once but do require hospitalization for 2-3 days.
Stage I disease Current recommendations for the treatment of patients with pathologic stage I disease include adjuvant pelvic irradiation for women with deep myometrial penetration, grade 2 or 3 histology, or evidence of vascular-space invasion . Data support the use of vaginal irradiation alone for women with more superficial tumors and low-grade histology.
Radiation doses are generally 45-50 Gy with standard fractionation. The technique should include multiple fields treated daily, with attempts to protect the small bowel. Complications from adjuvant pelvic irradiation are related to technique and the extent of lymphadenectomy.
Papillary serous histology The high rate of upper abdominal, pelvic, and vaginal recurrences in patients with uterine papillary serous cancers has led to the recommendation that they receive whole-abdominal irradiation with doses of up to 30 Gy and additional treatments to bring the pelvic dose to 50 Gy. A vaginal cylinder or colpostats may be used to boost the surface dose with 40 Gy. This treatment has resulted in a 5-year survival rate of 50%.
Stage II disease Patients whose endometrial cancer extends to the cervix usually are a heterogeneous group with differing histologic grades and varying degrees of cervical involvement, myometrial penetration, and nodal involvement. Similar outcomes with preoperative and postoperative irradiation suggest that initial surgical treatment with tailored postoperative irradiation is a reasonable approach.
Current treatment recommendations frequently include adjuvant pelvic irradiation to a dose of 45-50 Gy, in addition to insertion of a vaginal cylinder or colpostats to raise the total dose to the vaginal surface to 80-90 Gy. This treatment should result in a 5-year disease-free survival rate of 80%, with a locoregional control rate of 90%. Of course, outcome varies with the extent of myometrial penetration, degree of cervical involvement, and histologic grade of tumor.
Extensive cervical involvement Patients who have a large amount of cervical involvement that precludes initial hysterectomy are candidates for preoperative irradiation. A multiple-field technique is used to deliver a dose of 40-45 Gy with standard fractionation. A midline block may be inserted for the last 20 Gy to protect the rectum.
Intracavitary insertion with a standard Fletcher applicator, consisting of a uterine tandem and vaginal colpostats, delivers 20-25 Gy to point A (defined as 2 cm caudally and 2 cm laterally to the cervical os). Hysterectomy should follow in approximately 4-6 weeks. The expected 5-year disease-free survival rate for patients with disease that is this extensive is 70%-80%.
Stage III disease Women in whom endometrial cancer has spread outside the uterus (stage III disease) are at increased risk of death but have a wide spectrum of expected outcomes.
Isolated ovarian metastasis One subgroup found to have a relatively good prognosis is women with isolated ovarian metastasis. Five-year disease-free survival rates ranging from 60% to 82% have been reported in these women after hysterectomy and pelvic irradiation, depending on the histologic grade of tumor and the depth of myometrial penetration. Pelvic irradiation usually includes a dose of 45-50 Gy using standard fractionation. A vaginal boost with a cylinder or colpostats may add 30-35 Gy to the vaginal surface.
Extension to periaortic nodes In patients with disease extension to the periaortic nodes, the relapse-free survival rate after hysterectomy and adjuvant irradiation to the pelvis and peri-aortic regions is approximately 30%. It has been suggested that multiple sites of extrauterine involvement may be the single worst prognostic factor in these patients. Such women have a high recurrence rate and should be treated with extended-field irradiation. The standard treatment recommendation includes 45 to 50 Gy of radiation to a volume encompassing the pelvic and periaortic regions.
Whole-abdominal irradiation Because upper abdominal failures have been reported previously in patients with stage III disease, attention has focused on the role of whole-abdominal irradiation. Although subsets of patients have done well with whole-abdominal therapy, it is unclear whether this more aggressive therapy has any benefit over pelvic irradiation. The GOG is conducting a trial of whole-abdominal irradiation compared to chemotherapy. The Radiation Therapy Oncology Group (RTOG) is investigating the combination of pelvic radiation and chemotherapy in patients who are at risk for pelvic and distant sites of recurrence.
Definitive radiation treatment
For patients who are poor operative risks, definitive treatment with irradiation has produced excellent local control and survival rates. Such treatment is considered to be justified when the operative risk exceeds the 10%-15% uterine recurrence rate expected with radiation alone.
A more favorable outcome with definitive irradiation is related to low clinical tumor stage, less aggressive histologic variant, and use of brachytherapy for at least part of the treatment. Five-year disease-specific survival rates as high as 87%, 88%, and 49% have been reported in patients with stages I, II, and III or IV disease, respectively. Ten-year local control rates in patients with stage I, II, and III or IV disease were 84%, 87%, and 68%, respectively.
Treatment techniques with radiation alone for patients with early-stage disease and low-grade histology consist of uterine intracavitary insertions with Heyman or Simon capsules or an afterloading intrauterine tandem. Doses for intracavitary treatment range from 40 to 45 Gy prescribed to point A. Patients with more advanced disease, a large uterus, or aggressive histology generally receive both an intrauterine intracavitary insertion and external pelvic irradiation. External irradiation typically delivers 40-45 Gy to the pelvis, followed by intracavitary treatment that delivers 30-35 Gy to point A.
Complication rates Rates of serious complications attributable to irradiation range from 4%-5% with intracavitary treatment alone to 10%-15% with combined external and intracavitary irradiation.
Adjuvant systemic therapy
Only a few trials of adjuvant systemic therapy have been conducted in patients with early-stage endometrial cancer. At present, we believe that such therapy should not be recommended outside of the clinical trial setting.
Endocrine therapy Early uncontrolled trials suggested that progestin therapy might prolong progression-free interval and time to recurrence in patients with stage I and II lesions treated with initial surgery and irradiation. However, at least three subsequent randomized trials failed to show any survival benefit for progestins. A recent meta-analysis has demonstrated no advantage of adjuvant progestin therapy.
Chemotherapy To date, there is also no convincing benefit for adjuvant chemotherapy. The GOG compared doxorubicin with observation in 181 patients with high-risk, early-stage, endometrial carcinoma; at 5 years, there was no difference in recurrence rates. A subsequent phase II trial suggested a better outcome for 62 high-risk patients treated with adjuvant cyclophosphamide, Adriamycin, and cisplatin (CAP) when compared to historical controls. Further adjuvant trials utilizing combinations of doxorubicin and cisplatin (Platinol), possibly in combination with other active new agents, such as paclitaxel (Taxol), are warranted.
Treatment of recurrent or metastatic disease
Patterns of recurrence
Recurrent endometrial cancer is initially confined to the pelvis in 50% of patients. The major sites of distant metastasis are the abdominal cavity, liver, and lungs.
Following diagnosis and initial treatment, periodic evaluation, including history, physical examination, and pelvic examination, is recommended at 3- to 6-month intervals for the first 5 years and yearly thereafter. The use in asymptomatic patients of more extensive and more costly procedures, such as chest x-ray, CT imaging, and marker studies, is of questionable value and is unlikely to have a major impact on survival. Symptomatic patients should be evaluated as is appropriate.
After hysterectomy alone for endometrial cancer, approximately 50% of recurrences are pelvic and 50% are extrapelvic. It is clear that locoregional recurrences can develop in isolation, without distant metastasis, and salvage can be accomplished with high-dose irradiation.
Pelvic recurrences Five-year disease-specific survival rates as high as 51% have been reported in patients with isolated locoregional recurrences treated with radiation therapy. Factors that have an adverse impact on outcome are increased size of recurrence, young age, pelvic vs vaginal involvement, and treatment of the recurrence with external-beam irradiation only vs the addition of vaginal brachytherapy.
Radiation treatment for pelvic recurrence usually consists of external-beam irradiation with the addition of a brachytherapy boost that may include colpostats, a cylinder, interstitial needles, or seeds.
Treatment must be individualized based on the location and size of the recurrence and the boost method selected. The tolerance of normal tissues must be respected, but combined doses > 60 Gy have been associated with improved local control.
Extrapelvic recurrences For patients with recurrences outside the pelvis, irradiation is very effective in producing responses in localized symptomatic lesions. Therefore, radiation may be effective for palliation of such lesions in the lymph nodes, brain, or bones. Doses and protocols vary, depending on the site of recurrence.
Advanced disease with abdominal involvement Whole-abdominal radiation may also be effective palliative therapy in patients with advanced endometrial carcinoma involving the abdominal cavity. A randomized trial comparing radiation with chemotherapy in this setting (GOG protocol 122) is currently in progress.
Pelvic exenteration for pelvic recurrences after irradiation
Isolated pelvic central recurrence after irradiation is rare. Selected patients in whom it does occur may benefit from pelvic exenterative surgery. No large series have been published, but some long-term survivors have been reported.
Barakat et al reported on 44 patients who underwent pelvic exenteration for recurrent endometrial cancer at Memorial Sloan-Kettering Cancer Center between 1947 and 1994. Primary therapy usually consisted of total abdominal hysterectomy and bilateral salpingo-oophorectomy, with most patients receiving either preoperative or postoperative radiotherapy. Prior to exenteration, 10 (23%) of 44 patients had never received any form of radiotherapy. The median interval between initial surgery and exenteration was 28 months (range, 2-189 months).
Exenteration was total in 23 patients (52%), anterior in 20 patients (46%), and limited to posterior in 1 patient. One vascular injury led to the only intraoperative death. Major postoperative complications occurred in 35 patients (80%) and included intestinal/urinary tract fistulas, pelvic abscess, septicemia, pulmonary embolism, and cerebrovascular accident. Median survival for the entire group of patients was 7.36 months, with 9 patients (20%) achieving long-term survival (> 5 years). Although the long-term survival rate after this procedure is only 20%, it remains the only potentially curative option for the few patients with central recurrence of endometrial cancer who have not responded to standard surgery and radiation therapy.
Progestins produce complete and partial response rates of 15%-25% in patients with locoregional recurrence or distant metastases. The route, type, and dose of progestins do not appear to be related to response; hence, oral therapy is preferred.
In clinical practice, oral administration of 200 mg of medroxyprogesterone or 160 mg of megestrol acetate produces blood levels similar to those achieved with parenteral therapy (400-1,000 mg of medroxyprogesterone acetate IM weekly). A phase 0.1III trial conducted by the GOG comparing 200 mg to 1,000 mg of medroxyprogesterone acetate given orally daily found no differences between the two regimens, although it is noteworthy that the trends all favored the low-dose regimen (Table 3). Doses higher than 200 mg/d of medroxyprogesterone acetate, therefore, are clearly not warranted.
Several factors are predictive of a favorable response to progestin therapy. Patients with well-differentiated lesions are more likely to respond. A related observation is that a much higher percentage of grade 1 tumors have significant levels of estrogen and progesterone receptors; data show that lesions with higher receptor levels respond much more frequently to progestins. Response is almost always associated with better progression-free and overall survival.
The median time to progression for all patients treated with progestins is 3-4 months, and the median survival is 10 months.
Tamoxifen has been utilized in the treatment of endometrial carcinoma both in the salvage setting and as a first-line systemic treatment. The largest trial, a recent GOG study involving patients who had never received systemic therapy for endometrial carcinoma, reported a 13% response rate. For all recent studies, the response rates range from 0% to 13%. These data suggest that tamoxifen is not as active as progestins and is of little value as second-line therapy in patients who do not respond to progestins.
Some reports evaluate combined therapy with tamoxifen plus a progestin given sequentially in the hope that tamoxifen may increase progesterone receptor expression and, thus, increase the likelihood of response to progestins. None of the studies suggest an advantage for such an approach.
Other hormonal agents, such as gonadotropin-releasing hormone analogs and aminoglutethimide (Cytadren), have been studied to some extent in endometrial carcinoma. These agents do not appear to have sufficient activity to warrant further study.
Single agents Chemotherapy for advanced endometrial cancer focuses on three groups of agents with demonstrated activity
Active agents The anthracyclines studied include doxorubicin and epirubicin. In a total of 298 patients, doxorubicin produced a 27% response rate. Epirubicin, primarily in European studies including 27 patients, yielded a 26% response rate. Two platinum compounds have activity. Cisplatin, in 86 patients, elicited responses in 29%. Carboplatin (Paraplatin) produced a 31% response rate in 52 patients. Paclitaxel, in two studies involving 47 patients, yielded responses in 36%.
For all of these studies, the progression-free interval ranged from 4 to 7 months, with an overall survival range of 8-12 months. Approximately one-third of the responses were clinical complete responses with a substantially longer duration and better survival.
Agents with limited activity Other agents studied have included alkylating agents (cyclophosphamide [Cytoxan, Neosar], ifosfamide [Ifex]), altretamine (hexamethylmelamine [Hexalen]), fluorouracil (5-FU), methotrexate, mercaptopurine (Purinethol), vinblastine, etoposide, teniposide (Vumon), and mitoxantrone (Novantrone). All of these agents exhibited insufficient activity to warrant further study.
Combination regimens A number of phase II trials of combination regimens have been conducted. Ultimately, the relative merits of combination chemotherapy must be judged in the context of a randomized trial.
The only such study to date to show an advantage for a combination regimen is GOG protocol 107, which randomized 223 patients with advanced or recurrent disease to either doxorubicin (60 mg/m²) or doxorubicin (60 mg/m²) plus cisplatin (50 mg/m²), each given every 3 weeks. As shown in the combination demonstrated a superior response rate, a higher complete response rate, and a better progression-free interval. Overall survival did not differ significantly between the two treatments. These data establish doxorubicin plus cisplatin as the regimen of choice.
A recently completed GOG study (GOG protocol 163) compared doxorubicin plus cisplatin to doxorubicin plus paclitaxel based on the demonstrated activity of paclitaxel. Data from this study await maturation before analysis.
The current GOG study (GOG protocol 177) is comparing doxorubicin plus cisplatin to the three-drug combination of doxorubicin, paclitaxel, and cisplatin.
Chemotherapy plus progestins Combinations of chemotherapy plus progestins have been studied in a number of phase II trials. The only large, randomized trial evaluating this approach (GOG protocol 29) allocated patients with advanced or recurrent disease to receive either cyclophosphamide, doxorubicin, cisplatin, and megestrol acetate or melphalan (Alkeran), 5-FU, and megestrol acetate. In pilot studies, these two regimens had been reported to yield response rates of 75% and 94%, respectively. The randomized trial produced response rates of 36% and 38%, respectively, with no evident advantage of either combination over prior studies of single-agent doxorubicin with regard to response rate, progression-free interval, or overall survival. These results do not suggest any advantage for the combined use of chemotherapy and progestins.
Patients who have advanced or recurrent endometrial carcinoma should be considered for systemic therapy. Patients should first be offered the opportunity to participate in a clinical trial. Those who are ineligible or who choose not to participate should be treated according to current evidence.
Patients who have a grade 1 tumor and/or known progesterone-receptorpositive disease clearly benefit from treatment with progestins (response rate, 40%; median progression-free interval, 9 months; overall median survival, 14 months) and should be so treated. Those with a grade 2-3 tumor and/or known progesterone-receptornegative disease do not do well with progestin therapy (response rate, 12%; median progression-free interval, 3 months; overall median survival, 10 months) and should be considered for initial chemotherapy with doxorubicin plus cisplatin. The doxorubicin/cisplatin combination should also be considered for patients who do not respond to initial hormonal therapy.
Regimens that include both chemotherapy and hormonal therapy should not be considered outside of a clinical trial because of the lack of data supporting any advantage of these combinations. Likewise, sequential use of tamoxifen and progestins is not indicated because of the absence of enhanced efficacy.