Molecular genetic characterization of tamoxifen-associated endometrial cancer

OBJECTIVES: The non-steroidal, selective estrogen receptor modulator tamoxifen is currently the most extensively used hormonal agent for the prevention and treatment of estrogen receptor-positive breast cancer. Epidemiologic studies and clinical trials have shown an increased risk of endometrial cancer with tamoxifen exposure; however, few studies have examined these tumors on a molecular level. We sought to elucidate the molecular genetic alterations found in tamoxifen-associated endometrial cancer. METHODS: Twenty-nine breast cancer patients with a history of tamoxifen use who subsequently developed endometrial cancer were retrospectively identified and matched for endometrial histologic subtype and grade to 29 endometrial cancers from breast cancer patients never exposed to tamoxifen. Endometrial tumor tissue was microdissected and genomic DNA extracted for each case. Direct DNA sequencing of the most commonly mutated genes in sporadic endometrial cancer, PTEN, K-RAS, TP53, and CTNNB1, was performed in addition to microsatellite instability (MI) studies. Fisher's Exact Test was utilized for statistical analyses. RESULTS: Of 29 tamoxifen-associated cancers, 10 (34.5%) contained PTEN mutations compared to 13 (44.8%) of the non-tamoxifen-associated cancers (P = 0.59). All PTEN mutations were found in tumors with endometrioid histology, reflecting what is seen in sporadic endometrial cancers. Mutations of K-RAS, TP53, and microsatellite instability were present in similar frequencies between the two breast cancer groups, and moreover, these were similar to mutational frequencies found in sporadic endometrial cancer. CONCLUSION: Tamoxifen and non-tamoxifen-associated endometrial carcinomas arising in women with breast cancer contain similar genetic alterations to those of sporadic endometrial carcinomas.     

Development of endometrial cancer following radiation therapy for cervical carcinoma

The clinical and histologic findings in five cases of endometrial cancer, which developed following radiation therapy for squamous cell carcinoma of the cervix, are described. The mean age at endometrial cancer diagnosis was 69 years and average latency period from initial therapy to development of endometrial carcinoma was 13.4 years. For endometrial cancer, one patient had Stage Ib, one patient had Stage IIIa, two patients had Stage IIIc, and one patient did not undergo laparotomy. The histological types were carcinosarcoma in two patients, endometrioid adenocarcinoma, grade 3 in one patient, and clear cell carcinoma in one patient. All patients died of disease within 33 months of diagnosis. Endometrial cancers that develop after radiation treatment have a preponderance of high-risk histological subtypes, and consequently have a poor prognosis. Long-term follow-up should be mandatory for patients surviving radiation therapy for cervical cancer in order to detect and effectively treat second malignancies.     

Survival and prognostic factors in patients with synchronous ovarian and endometrial cancers and endometrial cancers metastatic to the ovaries

PURPOSE: To compare the survival and prognostic factors of patients with synchronous primary ovarian and endometrial cancers, and endometrial cancers metastatic to the ovaries. PATIENTS AND METHODS: Fifty-three patients with synchronous primary ovarian and endometrial cancer and 64 patients with endometrial cancer metastatic to the ovaries were evaluated. RESULTS: Mean follow-up time was 47.2 months (18-170 months). There was no statistical difference in age, gravidity and parity between the two groups. Abnormal vaginal bleeding was the most common symptom in both groups. All patients were subjected to a surgical staging procedure. Overall survival of the synchronous group was significantly higher than that of the metastatic group (98 +/- 12 vs 59 +/- 6 months; p = 0.048). The significant prognostic factors for synchronous cancers after multivariate analysis were age, stage of ovarian cancer, grade of endometrial cancer, and adjuvant therapy status. CONCLUSION: Patients with synchronous ovarian and endometrial cancers appear to have a good prognosis and should undergo primary surgical staging since the stage of tumors is a significant prognostic factor.     

Para-aortic node biopsy in cervical and endometrial cancers: does it affect survival?

Two hundred ten patients with endometrial and cervical carcinoma had para-aortic node biopsies. Nineteen of the 210 patients (9.0%) had positive para-aortic nodes. These 19 patients received pelvic irradiation, and 18 patients received para-aortic irradiation. The incidence of para-aortic nodal involvement in cervical carcinoma was directly related to the stage of the disease. Eleven of the 12 patients with cervical carcinoma and positive para-aortic nodes received both pelvic and para-aortic irradiation. Three of these patients are alive without disease, resulting in a survival rate of 25%. These patients are surviving for 16, 30, and 41 months. The incidence of positive para-aortic nodes in endometrial adenocarcinoma was related to the uterine length and the histologic grade. The survival rate for patients with endometrial adenocarcinoma and positive para-aortic nodes in this study was 57.1%. Four patients have survived for 1, 30, 60, and 71 months. There were no surgical deaths or radiation therapy complications directly attributable to para-aortic biopsy or irradiation.     

Abnormal cervical cytology: a risk factor for endometrial cancer recurrence

The objective of this study was to evaluate the relationship between cervical cytology, histologic type, and risk of endometrial cancer recurrence. We performed a retrospective study of patients undergoing surgery for endometrial carcinoma. Risk factors for recurrence including histology, tumor grade, nodal status, myometrial invasion, peritoneal washings, stage, and cervical cytology were assessed. Abnormal cervical cytology was defined as the presence of any endometrial cells on Pap smear. Papillary serous and clear cell carcinomas were considered high-risk histologies. Univariate and multivariate analyses of risk factors for recurrence were performed. Thirty-nine (9%) patients developed recurrent endometrial cancer. More patients with abnormal Pap smears recurred (12% versus 4%, P < 0.05). For endometrioid adenocarcinoma, abnormal cervical cytology occurred in 61% and 7% recurred, while with high-risk histologies, 84% had abnormal cervical cytology and 19% recurred (P < 0.05). Other significant predictors of recurrence on univariate analysis were myometrial invasion, nodal status, washings, stage, and histology. On multivariate analysis, only nodal status remained a significant predictor of recurrence. Abnormal cervical cytology is associated with increased risk of endometrial cancer recurrence. Abnormal cervical cytology occurs more frequently in high-risk histologies, which are known to have a higher risk of recurrence. On multivariate analysis, only nodal spread remains a significant predictor of recurrence.     

p53 staining as a prognostic indicator in endometrial carcinoma.

OBJECTIVE: To investigate p53 expression in endometrial cancer and its significance as a prognostic indicator. METHODS: Thirty-five consecutively surgically treated patients with endometrial cancer had their p53 expression studied by immunoperoxidase staining and quantified by lighted microscopic evaluation of the staining pattern. The determination of mean percentage of p53 expression was compared to prognostic indicators of endometrial cancer. RESULTS: p53 staining was detected in 20 of the 35 cases of endometrial carcinoma. Eleven of the 21 endometrioid tumors stained positive, while 9 out of 14 tumors with more aggressive histology stained positive for p53. If the grade I and II patients were taken into account as a whole, there was a statistically significant correlation (p<0.001) between the grade I and II patients and the grade III patients. The difference was statistically significant between stage I and III (p<0.05). The difference between lymphovascular space invasion and no lymphovascular invasion and p53 positivity was statistically significant (p<0.05). CONCLUSION: p53 expression is more common in more aggressive histologic subtypes than in endometrioid adenocarcinomas. Strong expression of p53 correlates with advanced stage and high grade and is detected more frequently in endometrial cancers with lymphovascular invasion.     

Absence of PTEN repeat tract mutation in endometrial cancers with microsatellite instability

OBJECTIVE: PTEN, a tumor suppressor gene shown to be frequently mutated in endometrial cancers, has been suggested to be a target of microsatellite instability (MSI)-driven mutagenesis. We set out to investigate the relationship between MSI and PTEN mutation in a large series of primary endometrial carcinomas. METHODS: Thirty-nine MSI-positive endometrial cancers were evaluated by single-strand conformational variant analysis and direct sequencing to screen all nine PTEN exons for mutation. RESULTS: Fifteen specimens (38%) demonstrated 16 PTEN mutations. We observed only one alteration in the poly-adenine repeat of exon 8 that is suggested to be a target for mutation in endometrial cancers with MSI. Seven of 16 (44%) mutations in our series were deletions of >/=3 bp, a class of mutation not usually associated with tumors with defective DNA mismatch repair. To determine the significance of this high frequency of deletion, 26 additional endometrial cancers without MSI were matched with the 39 MSI-positive cancers for the prognostic factors of tumor histology, stage, grade, and patient race. The MSI-positive tumors had a significantly higher frequency of deletions involving >/=3 bp when compared with the MSI-negative group (5/11 versus 0/10, P = 0.035). CONCLUSIONS: Repeat tract mutation in PTEN is an uncommon event in MSI-positive cancers. Deletion of >/=3 bp in this gene is more common in MSI-positive cancers when compared with tumors without MSI.     

A semiautomated test for microsatellite instability and its significance for the prognosis of sporadic endometrial cancer in northern Norway.

Archival histologic material from 105 women (median age 62 years) treated for endometrial cancer was investigated for the replication error phenotype indicated by the observation of widespread microsatellite instability (MSI). Polymerase chain reaction (PCR) of DNA isolated from paraffin-embedded tissue was analyzed for MSI using six microsatellite loci with a fluorescent-based detection system. Flow cytometry and morphometric investigation were performed in the same material for each of the patients. Twenty percent (21 of 105) of screened endometrial cancers were found to have high MSI at two or more of the loci tested. The mean detection frequency per marker was highest in the dinucleotide repeat sequence, D2S123, and the mononucleotide repeat sequences amplified by Bat 25 and Bat 26. Death from endometrial cancer was not related to the occurrence of MSI (p=0.6). There was no significant association between MSI and FIGO stage (p=0.5), myometrial invasion depth (p=0.8), histological grade (p=0.3), or vessel invasion (p=0.5). There were, however, more MSI cases among the group of diploid cases compared with the aneuploid and tetraploid group. MSI is not a valuable prognosticator for survival of sporadic endometrial cancer, and diploid cases are significantly more often MSI positive than aneuploid and tetraploid cases.     

Management of endometrial cancer with suspected cervical involvement

The 1989 International Federation of Gynecology and Obstetrics (FIGO) staging system for endometrial cancer cells for operative assessment of the extent of uterine disease, grade, and sites of metastasis before assigning a stage to the cancer. In the current study, 70 endometrial cancer patients with suspected cervical involvement based on a positive endocervical curettage or punch biopsy were treated with initial surgery followed by tailored radiation or chemotherapy. Only 37% of the patients had operative findings consistent with the preoperative suspicion of stage II disease. Postoperative therapy was determined by the extent of cervical involvement, depth of myometrial invasion, cell type, tumor grade, and the presence and location of extra-uterine disease. Based upon these parameters, 21 patients were believed to have low risk for pelvic recurrence and received no adjuvant therapy (90% 5-year survival); 38 patients received postoperative pelvic radiation because of high-risk factors for pelvic recurrence or pelvic nodal involvement (65% 5-year survival); and 11 patients received chemotherapy and/or extended radiation because of extrapelvic disease (no 5-year survivors). The approach outlined supports initial surgery for cases of endometrial cancer with suspected cervical involvement. This approach permits accurate surgical staging under the new FIGO system, avoids radiotherapy in many patients whose disease is less extensive than suspected preoperatively, and can accomplish good local control with limited morbidity.     

Pathological findings in early-stage endometrial cancer

OBJECTIVE: The aim of this study was to assess the pathological characteristics of early-stage endometrial cancer, with regard to endometrioid versus serous papillary adenocarcinoma. METHODS: Sixty-six cases of early-stage endometrial carcinoma were classified into two groups: group I--36 cases of endometrioid endometrial cancer, staged IA-IB and graded G1-G2; group II--30 cases of Stage I serous papillary endometrial cancer. The pathological characteristics compared between the two groups included features such as tumor location in the uterine cavity, tumor focality, lymphovascular invasion, as well as the status of the uninvolved endometrium, adjacent to the tumor. Patient clinical characteristics were obtained from the medical records. RESULTS: Significantly more patients with endometrioid endometrial cancer were premenopausal (p < 0.0001), obese (p < 0.02), had hypertension (p < 0.00001) and familial cancer (p < 0.0001). On the other hand, significantly more patients with serous papillary cancer had another primary malignancy (p < 0.001). Considering the pathological characteristics, 75% of endometrioid as compared with 6.7% of serous papillary cancer cases were found in the upper uterine segment only (p < 0.0001). Multifocality was observed in 16.7% of endometrioid as compared with 100% of serous papillary cancer cases (p < 0.0001). Lymphovascular space invasion was absent in all cases of endometrioid cancer, while present in 90% of serous papillary cancer cases (p < 0.0001). Seventy-five percent of endometrioid and 100% of serous papillary cancer cases were associated with an atrophic endometrium. CONCLUSION: The clinical and pathological features of early-stage endometrial cancer differ according to the histological type of the cancer. The majority of endometrioid cancers are probably associated with an atrophic or normally cycling endometrium, and not with endometrial hyperplasia.     

Tumor cytotoxicity of peritoneal macrophages and peripheral blood monocytes from patients with ovarian, endometrial, and cervical cancer

Abstract. Wilbanks D, Ahn M-C, Beck DA, Braun DP. Tumor cytotoxicity of peritoneal macrophages and peripheral blood monocytes from patients with ovarian, endometrial, and cervical cancer.
The purpose of this study was to compare the cytotoxic capacity of peritoneal macrophages (PM) and peripheral blood monocytes (PBM) from patients with ovarian, endometrial, and cervical cancers after in vitro activation with gamma interferon (IFN-γ) and lipopolysaccharide (LPS). Peritoneal macrophages were obtained from ascites or peritoneal washings and peripheral blood monocytes via peripheral venipuncture from 58 patients: 17 with ovarian, 19 with endometrial, and 10 with cervical cancers. PBM and PM from 12 patients with nonmalignant gynecologic conditions served as controls. Cytotoxicity was assessed by the ability of PBM and PM to lyze Cr⁵¹-labeled Chang hepatoma cells. Activated peripheral blood monocytes of ovarian and endometrial cancer patients and peritoneal macrophages from ovarian cancer patients were significantly more cytotoxic than those from nonactivated controls. Activated PBM and PM from cervical cancer and PM from endometrial cancer did not demonstrate increased cytotoxicity compared to nonactivated controls. There was no significant correlation of the cytotoxicity with grade, stage, differentiation or age of the cancers. These in vitro data would suggest that ovarian cancer and possibly endometrial cancer should receive further evaluation and consideration of cytokine-based and/or adoptive cellular immunotherapy.     

Surgical staging for patients presenting with grade 1 endometrial carcinoma

OBJECTIVE: To examine the impact of surgical staging of patients presenting with grade 1 endometrial cancer. METHODS: The charts of all patients who presented for surgery for endometrial cancer between March 1997 and July 2003 were analyzed for demographic data, final tumor histology, grade, stage, and complications. RESULTS: A total of 349 patients underwent surgical management for endometrial cancer. Preoperatively, 181 (52%) were identified with grade 1 disease, with a mean age of 61 years (range 27-89). Surgical staging (pelvic +/- para-aortic lymphadenectomy) was performed in 82% of cases and was omitted only in cases when disease was apparently confined to the endometrium and surgical risk was high. In staged patients, 3.2% had severe surgical complications. There were 2 perioperative mortalities (1 pulmonary emboli and 1 myocardial infarct). In comparison of pre- and postoperative histology, 19% of patients were upgraded, with 15% grade 2, 0.5% grade 3, 2.5% serous or clear cell, and 1% mixed mesodermal tumor. Lymph node metastases were found in 3.9% of patients presenting with grade 1 endometrial cancer, and 10.5% had extrauterine spread (> IIb). High-risk uterine features, including myometrial invasion more than 1/2, grade 3 lesions, high-risk histologic variants, and/or cervical involvement, were found in 26% of the patients. No patients with stage Ia-IIb endometrioid cancer received adjuvant teletherapy or chemotherapy. Four patients with low-risk uterine features were found to have extrauterine disease. Twelve percent of patients received adjuvant therapy, and 17% avoided teletherapy and/or chemotherapy based on surgical staging. CONCLUSION: Surgical staging in patients presenting with grade 1 endometrial cancer significantly impacted postoperative treatment decisions in 29% of patients. Omitting lymphadenectomy in patients presenting with grade 1 endometrial cancer may lead to inappropriate postoperative management.     

Favorable survival associated with microsatellite instability in endometrioid endometrial cancers

OBJECTIVE: To determine whether microsatellite instability in endometrioid endometrial cancer is associated with favorable survival. METHODS: Microsatellite instability analysis was performed in 131 patients with endometrioid endometrial cancer using three polymorphic markers in paired cancer and normal DNA. Logistic regression and multivariable analyses calculated the relation between microsatellite instability, clinical features, and survival. RESULTS: Microsatellite instability was detected in 29 of 131 (22%) endometrioid endometrial cancers. There was no correlation between microsatellite instability and age, race, grade, stage, or depth of myometrial invasion. Microsatellite instability was associated with better survival in univariate and multivariable analyses after controlling for confounding influences (P =.03). The 5-year survival rate of those with microsatellite instability was 77% (95% confidence interval 55%, 90%) compared with only 48% (95% confidence interval 39%, 57%) in other cases. Microsatellite instability was associated with other molecular features that predict favorable outcome including PTEN mutation (P =.002) and the absence of p53 overexpression (P =.01). CONCLUSION: Microsatellite instability is a molecular alteration associated with favorable outcome in endometrioid endometrial cancers, even when accounting for other prognostic factors. This association might be explained by the finding that the pathway of molecular carcinogenesis characterized by loss of DNA mismatch repair favors alteration of genes that result in a less virulent clinical phenotype.     

Clinico-pathologic comparison of type II endometrial cancers based on tamoxifen exposure

Objective

To compare clinico-pathologic variables and protein expression of potential regulatory components in patients who develop type II endometrial cancer with and without antecedent tamoxifen.

Methods

Clinico-pathologic variables were compared for all surgically staged patients (2000-2008) with grade 3 endometrioid, papillary serous, clear cell, and carcinosarcoma of the uterus based on tamoxifen exposure [Tam (+) vs. Tam (−)]. Overall survival was analyzed using a multivariable Cox regression model and Kaplan-Meier estimates. Protein expression of ERα, PR, mTOR, p-mTOR, IGF-1R, EGFR, VEGF and HER-2/neu was compared by immunohistochemistry using a semiquantitative scoring system.

Results

Of 115 patients with high grade endometrial cancers, 15 received tamoxifen. These patients were older at diagnosis than Tam (−) patients. A higher percentage of Tam (+) patients had carcinosarcoma compared to Tam (−) patients (60% vs. 30%, P = 0.038). Overall survival for Tam (+) patients was shorter than Tam (−) patients (16.6 vs. 32.2 months, P = 0.004). The hazard ratio for death for Tam (+) patients was 2.53 (P = 0.014), controlling for age and stage. Intensity and extent of staining were similar for ERα, PR, VEGF, EGFR, p-mTOR and HER-2/neu. The average expression score for IGF-1R and mTOR in the Tam (+) group was significantly higher than the Tam (−) group: 10.3 vs 7.0, P = 0.001 and 6.0 vs 3.1, P = 0.029, respectively.

Conclusion

There are differences in the biology of type II endometrial cancers that develop in women with prior tamoxifen exposure. Tamoxifen associated cancers show higher expression of IGF-1R and mTOR, which should be further investigated.     

Treatment of stage II endometrial carcinoma

The optimal management of stage II carcinoma of the endometrium remains to be established. We reviewed our experience in treating 42 patients with stage II endometrial cancer by surgery, radiation, or combined radiation and surgery at the Hospital of the University of Pennsylvania. The overall 5-year survival was 47.6%. The 5-year survivals of patients treated by surgery only, radiation only, or combination radiation and surgery were 68.5, 36.5, and 46.1%, respectively, which were not significantly different. Histologic grade was found a significant prognostic factor but type of cervical involvement was not. Major complication rates were similar in each treatment group. We conclude that the majority of patients with stage II endometrial carcinoma are best treated by combination radiation and surgery, but in a select subset of patients, radical hysterectomy and lymphadenectomy constitute a reasonable treatment option.     

冰冻病理诊断预测子宫内膜癌淋巴转移的临床价值研究

目的:研究影响子宫内膜癌患者淋巴结转移的因素,评价术中冰冻病理预测淋巴结转移的作用。方法:回顾分析1996年7月至2008年1月在上海交通大学医学院附属仁济医院和2008年9月至2011年9月在同济大学附属第一妇婴保健院收治的共389例子宫内膜癌患者的临床资料,195例患者实施了盆腔淋巴结切除,其中43例同时行腹主动脉旁淋巴结切除。分析患者淋巴结转移的临床相关因素,评价冰冻病理结果在预测淋巴结转移中的价值。结果:盆腔淋巴结转移率为12.8%(25/195),腹主动脉旁淋巴结转移率为11.6%(5/43)。深肌层浸润(P<0.001)、宫颈累及(P<0.001)、ER阴性(P=0.001)与盆腔淋巴结转移显著相关。肿瘤细胞级别升高、病理类型(Ⅰ型、Ⅱ型)与盆腔淋巴结转移无显著相关性。低风险子宫内膜癌(排除G3和肌层深度≥1/2)患者的盆腔淋巴转移率为4.5%(3/67)。按冰冻结果制定4种预测模型,G1+限于内膜组,淋巴结阳性率为0;G1+<1/2肌层组,盆腔和腹主淋巴结阳性率均为2.4%;G2+<1/2肌层组,盆腔和腹主淋巴结阳性率分别为4.8%、0;未发现G2+限于内膜的病例。淋巴结切除组的生存率高于未切除组(79.5%vs 75.9%),但无统计学差异(P=0.086)。结论:冰冻病理用于预测淋巴结转移的作用有限,建议对除G1限于内膜的子宫内膜样腺癌患者,其余均应实施全面的分期手术。     

Preoperative radiotherapy for early endometrial carcinoma

A retrospective study was undertaken to compare the use of one versus two preoperative radium systems for early endometrial carcinoma. The charts of 73 patients treated between 1977 and 1980 were reviewed. No difference was noted between the two groups when compared for stage, grade, depth of myometrial invasion, and histologic type of tumor. One of thirty-eight (2.6%) patients in the one-radium group developed an isolated central recurrence; there were no central recurrences in the two-radium group. Total duration of therapy and total hospitalization for the one-radium versus the two-radium group were 17.6 and 15.3 days versus 77.0 and 17.3 days, respectively. Follow-up ranged from 48 to 84 months. Corrected survival figures are comparable to 94.6% for the one-radium group versus 100% for the two-radium group. These data suggest comparable effectiveness and morbidity between the two treatment regimens, with the single-radium application more efficient and cost effective.     

Shed membrane fragment-associated markers for endometrial and ovarian cancers

OBJECTIVE: The objective of this study was to assess circulating tumor-derived membrane fragments (MFs) and specific proteins associated with them as diagnostic and prognostic markers for ovarian and endometrial cancers. METHODS: The level of shed tumor-derived plasma MFs was analyzed chromatographically on high exclusion limit agarose-based gels, using sera from non-cancer-bearing female controls (n = 50), women with ovarian disease (benign, n = 43, and stages III and IV ovarian cancer, n = 62), and women with early (n = 10) and late (n = 12) stage endometrial cancers. The presence of specific proteins on MFs associated with development and progression of cancer was examined by Western immunoblot, while the association of proteolytic enzymes with MFs was analyzed by zymography. RESULTS: Shed MFs were demonstrated in the circulation of women with both ovarian (4.12 plus minus 1.46 mg/ml) and endometrial cancers (2.53 +/- 0.34 mg/ml in women with stage I and 4.51 +/- 1.33 mg/ml with late stage); however, they were not demonstrated in control sera or in sera from women with benign disease. In endometrial cancer, the level of MFs correlated with the tumor's progression. Specific proteins, including MMP-2, MMP-9, and Fas ligand (FasL), were present on MFs from both endometrial and ovarian cancer sera. However, FasL (3.2-fold) and MMP-2 and -9 (5.9x and 7.5x, respectively) levels were significantly elevated on MFs from late stage cancer. CONCLUSION: This study demonstrates the unique elevation of circulating MFs in ovarian and endometrial cancer patients. The levels of specific MF-associated proteins differentiate between early and late stage endometrial cancers and benign versus malignant ovarian disease.     

Cost-effectiveness analysis of strategies for the surgical management of grade 1 endometrial adenocarcinoma

OBJECTIVE: To estimate the costs and outcomes of various strategies used for the management of grade 1 endometrial cancer. METHODS: A cost-effectiveness analysis compared three strategies for the management of grade 1 endometrial cancer: 1) surgical staging in all patients (including hysterectomy and lymphadenectomy); 2) frozen section following hysterectomy with surgical staging based on the results of tumor grade and depth of myometrial invasion; and 3) hysterectomy without surgical staging (no staging). Surgical probabilities and recurrence rates were estimated from published data. Actual payer costs of surgery, radiation therapy, and chemotherapy were estimated for each strategy. Cost-effectiveness ratios were estimated for each strategy. Sensitivity analyses evaluated the costs of radiation and survival estimates used in the model. RESULTS: For the estimated 10,000 women diagnosed annually with grade 1 endometrial cancer in the United States, the annual cost of surgical staging is $240.4 million, compared with $252.4 million for frozen section and $255.8 million for no staging. Five-year disease-free survival for surgical staging is 87.9%, compared with 87.3% for frozen section and 86.7% for no staging. This translates into a lower cost-effectiveness ratio for surgical staging ($27,337) compared with frozen section ($28,913) or no staging ($29,513). Surgical staging yielded 64 additional disease-free patients per 10,000 patients compared with frozen section and 126 additional disease-free patients compared with no staging. Use of adjuvant radiation therapy was the lowest in the surgical staging strategy (13%). CONCLUSION: Surgical staging of all patients with grade 1 endometrial cancer is the most cost-effective strategy and decreases the use of radiation therapy without negatively impacting survival.