Endometrial cancer remains the most common gynecologic malignancy. The American Cancer Society estimated that almost 40,000 new cases were diagnosed, and approximately 7,500 deaths resulted from the disease in 2008.1 Surgical staging of endometrial cancer provides pathologic and prognostic information that can help guide management decisions with regard to postoperative adjuvant therapy. Stage IIIA endometrial cancer encompasses a heterogeneous group of tumors, including women with extrauterine disease manifested by adnexal metastases, uterine serosal involvement, or positive cytology alone.
Although extrauterine spread of endometrial cancer is an accepted negative prognostic factor, the importance of positive peritoneal washings is less certain. Prior studies have recognized the possible differences in clinical behavior and outcomes in stage IIIA endometrial cancer by subcategorizing the stage into IIIA1, extrauterine disease reflected by malignant cytology only, and IIIA2, extrauterine disease manifested by adnexal and serosal involvement. While some studies have suggested that positive cytology alone is associated with decreased survival, others have reported that positive cytology alone is of little clinical significance.2–13 Many of the prior studies of stage IIIA endometrial cancer have been limited by small sample size, the inclusion of nonendometrioid histologic subtypes, and lack of comprehensive staging data. The confusion surrounding the effect of positive washings has led to widely disparate recommendations for the treatment of stage IIIA1 carcinomas.14
The goal of our study was to examine the natural history of stage IIIA endometrioid endometrial adenocarcinomas in a large cohort. Specifically, we compared the outcomes of stage IIIA1 and IIIA2 tumors. To further explore the significance of positive cytology alone, we compared the outcomes of patients with stage IIIA1 tumors to women with stage IB and IC carcinomas.
MATERIALS AND METHODS
Data from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) database were used. Surveillance, Epidemiology, and End Results15 is a population-based cancer registry that includes approximately 26% of the United States population. Surveillance, Epidemiology, and End Results is composed of a number of geographically distinct tumor registries. The demographic characteristics of the SEER registries have been previously characterized.16,17 Data from SEER 17 registries were used. Exemption from the Columbia University Institutional Review Board review was obtained.
Women with tumors of the uterine corpus classified as endometrioid carcinoma (8.380/3) and adenocarcinoma not otherwise specified (8.140/3) treated between 1998 and 2004 were examined. Clinical and pathologic data including age at diagnosis (younger than 40, 40–65, and older than 65 years), race (white, African American, Asian, or other), marital status, tumor grade, receipt of radiation, and performance of lymphadenectomy were collected. Year of diagnosis was stratified as 1998–2001 or 2002–2004 for analysis. Receipt of radiation was classified as vaginal brachytherapy, external beam (with or without vaginal brachytherapy), none, or unknown. Patients were categorized based on the geographic area of residence within the United States at the time of diagnosis: central (Detroit, Iowa, Kentucky, Louisiana), eastern (Connecticut, New Jersey, Atlanta, rural Georgia), and western (Alaska, California, Hawaii, Los Angeles, New Mexico, San Francisco, San Jose, Seattle, Utah). Staging information was derived from the American Joint Cancer Committee staging information and recorded extent of disease codes. Primary analysis focused on women with stage IIIA1 (positive cytology) and IIIA2 (uterine serosal or adnexal metastasis). Patients with stage IB and IC tumors were also examined.
The vital status of each patient was recorded. Survival was calculated as the number of months from cancer diagnosis to date of death. Patients who were alive at last follow-up were censored. Both overall and cancer-specific survival was calculated for all patients.
Patients with stage IIIA1 tumors were compared with those with IIIA2 lesions. A separate analysis comparing stage IIIA1 cancers to women with IB and IC tumors was also performed. Frequency distributions between categorical variables were compared using χ2 test. Cox proportional hazards models were developed to examine survival. In Cox proportional hazards analyses, we modeled the cancer-specific and overall mortality hazard ratios comparing patients who IIIA1 and IIIA2, controlling for the other predictive variables. Separate Cox models were generated that included only patients with IIIA1, IB, and IC tumors. Separate analyses were performed that included only patients who underwent lymphadenectomy. Kaplan-Meier curves were generated, and overall and cancer-specific survival was examined using the log rank test. All hypothesis tests were two-tailed. Statistical analysis was conducted using R Version 188.8.131.52
A total of 1,257 patients with stage IIIA endometrioid endometrial adenocarcinoma were identified. The cohort included 485 (39%) women with IIIA1 tumors and 772 (61%) with IIIA2 lesions. The clinical characteristics of the population are displayed in Table 1. Subjects with IIIA1 and IIIA2 tumors were similar with respect to age at diagnosis, racial distribution, and marital status. Patients with IIIA1 carcinomas were more likely to have lower-grade histology than those with IIIA2 tumors; 27.4% compared with 17.4% had grade 1 tumors, whereas 29.9% compared with 36.9% had grade 3 tumors, respectively (P<.001). Those with IIIA1 tumors were also less likely to have undergone adjuvant radiation therapy (49.9% compared with 59.1%, P<.001) and more likely to have undergone lymphadenectomy (64.3% compared with 57.0%, P=.01).
In a Cox proportional hazards model incorporating other known prognostic variables, there was no difference in either cancer specific (hazard ratio [HR] 0.95, 95% CI 0.72–1.24) or overall (HR 1.06, 95% CI 0.85–1.32) survival between those with IIIA1 and IIIA2 tumors (Table 2). Independent predictors of decreased survival for stage III endometrial cancer included advanced age, residence in the eastern or central United States, and high tumor grade. Performance of lymphadenectomy (HR 0.48, 95% CI 0.39–0.59) was associated with improved survival. Although external beam radiotherapy had no effect on cancer-specific survival (HR 0.87, 95% CI 0.67–1.13), it did improve overall survival (HR 0.75, 95% CI 0.60–0.93). Similar findings were noted in a Kaplan-Meier analysis; there was no difference in either cancer-specific (P=.46) or overall survival (P=.70) between stage IIIA1 and stage IIIA2 (Figs. 1 and 2).
To eliminate potential bias from women with occult nodal disease, we performed a subgroup analysis in the 752 (60%) patients who underwent lymphadenectomy (Table 2). Again, there was no difference in either cancer specific (HR 0.78, 95% CI 0.52–1.16) or overall (HR 0.86, 95% CI 0.61–1.19) survival between stage IIIA1 and IIIA2 patients in a proportional hazards model. Stage, grade, area of residence, and race remained as predictors of survival. In a Kaplan-Meier analysis, there was no difference in either cancer-specific (P=.53) or overall survival (P=.85) between stage IIIA1 and IIIA2.
The outcome of patients with stage IIIA1 tumors was then compared with that of women with stage IB and IC lesions. In a multivariable Cox model, mortality was higher for IIIA1 tumors than both stage IB (HR 0.24, 95% CI 0.20–0.29) and stage IC (HR 0.46, 95% CI 0.38–0.55) cancers. Likewise, when the analysis was limited to only women who had undergone lymphadenectomy, survival was inferior for IIIA1 lesions compared with both IB (HR 0.27, 95% CI 0.21–0.36) and IC (HR 0.51, 95% CI 0.38–0.66) tumors. In a Kaplan-Meier analysis, cancer-specific and overall survival were inferior for women with IIIA1 tumors compared with stage IB and IC tumors (P<.001 for all comparisons) (Fig. 3).
Five-year survival was then examined. Five-year survival for stage IIIA1 disease (67%; 95% CI 61–72%) was inferior to that of both stage IB (90%; 95% CI 89–90%) and stage IC (77%; 95% CI 75–79%) tumors (Table 3). After stratification by grade, 5-year survival for stage IIIA1 tumors was worse for all grades when compared with stage IB disease. Five-year survival overlapped for stage IC and IIIA1 patients with grade 1 and 2 tumors. However, for women with grade 3 cancers, 5-year survival was lower for stage IIIA1 (53%; 95% CI 44–62%) than stage IC (68%; 95% CI 63–72%). Similar trends were noted when stage IIIA2 patients were compared with women with IC lesions.
Our findings suggest that women with stage IIIA endometrial cancer, even those with only positive cytology, are at significant risk to die from their disease. Patients with stage IIIA1 tumors seem to have an outcome that is similar to that of women with serosal or adnexal metastases. The outcome of stage IIIA1 endometrial cancer is inferior to that of tumors confined to the corpus (stage IB and IC).
Prior investigators have postulated that stage IIIA endometrial cancer is composed of two distinct groups with unique natural histories. The theory that patients with positive washings as the only manifestation of extrauterine disease have a better prognosis than those who have “true” extrauterine metastases arises from the premise that cancer cells found in peritoneal washings may have lower malignant potential and thus less biologic ability to result in local recurrence or distant metastases. Studies comparing the outcomes of IIIA1 and IIIA2 tumors have been in conflict; although some studies suggest no difference in survival, others have found that patients with adnexal metastases or serosal involvement have a worse outcome.2–5,19 Preyer et al4 demonstrated that patients with IIIA1 tumors had an improved survival compared with IIIA2 tumors. In an examination of 51 patients with stage IIIA endometrial cancer, Mariani and co-workers2 noted that whereas recurrence-free survival was lower in patients with IIIA2 tumors, 5-year disease-related survival was similar for IIIA1 (86%) and IIIA2 (93%) lesions. Our findings suggest that the outcome is similar for women with IIIA1 and IIIA2 tumors. In our cohort, 5-year survival was 67% for IIIA1 lesions and 55% for IIIA2 tumors.
Many clinicians have hypothesized that positive cytology alone is of little prognostic significance and that outcomes are primarily based on uterine risk factors such as grade, depth of invasion, and lymphvascular space invasion. As such, the outcomes of patients with stage IIIA1 tumors should be similar to that of patients with stage I disease with matched uterine risk factors. Studies comparing stage IIIA1 to stage I lesions have also been in conflict.3,6,19–21 In a multivariable analysis, Havrilesky and coworkers3 noted that positive cytology was associated with a twofold increase in death. In contrast, in an analysis of more than 300 patients recorded in the Geneva Cancer registry,19 outcomes were similar for stage I and IIIA1 patients. Five-year disease-specific survival was 92% for stage I lesions and 91% for stage IIIA1 tumors. This study is limited by the fact that only 2% of patients underwent lymphadenectomy. Our findings clearly suggest that patients with stage IIIA1 tumors fare worse than women with IB or IC neoplasms even when analyzing only those patients who underwent lymphadenectomy.
Despite the significant mortality associated with stage IIIA endometrial cancer, there is no consensus on the optimal adjuvant treatment of the disease.22 Treatment options vary widely and include observation, vaginal brachytherapy, whole pelvic radiation, whole abdominal radiotherapy, intraperitoneal P32, systemic hormonal therapy, cytotoxic chemotherapy, or combinations of the above. In a survey of practice patterns of the Society of Gynecologic Oncologists in 2005, there was a clear difference in recommendations for adjuvant therapy for patients with IIIA1 compared with IIIA2 disease. Adjuvant therapy was recommended only 46% of the time for IIIA1 grade 1 tumors and 62% of the time for IIIA1 grade 2 tumors. In contrast, adjuvant therapy was recommended 99% of the time for those with IIIA2 disease.14 Further trials to examine optimal management strategies for this heterogenous group of patients are clearly warranted.
Our study benefits from the inclusion of a very large cohort of patients with IIIA tumors, all of which were endometrioid histology. However, a number of limitations must be acknowledged. Although SEER data provide a representative sample of the U.S. population, as with any registry study, centralized pathology review is not available. Surveillance, Epidemiology, and End Results lacks important pathologic data, including lymphvascular space invasion and depth of myometrial invasion for patients with stage IIIA tumors. To correct for the fact that not all patients underwent lymphadenectomy, we performed a separate analysis that included only those patients who underwent nodal assessment. Finally, although data on radiation use are available, other treatment modalities such as chemotherapy and hormonal therapy are not recorded in the SEER database.
Our study suggests that women with stage IIIA endometrial cancer are at significant risk of mortality from their disease. These findings support the current International Federation of Gynecology and Obstetrics staging criteria and demonstrate that all patients with stage IIIA disease, including those with extrauterine cancer cells manifested only in positive peritoneal washings, may benefit from adjuvant therapy. Future research focusing on sites of recurrence in patients with IIIA endometrial cancer and outcomes after different adjuvant therapies are employed may help standardize postsurgical treatment recommendations and improve outcomes.
1. Jemal A, Siegel R, Ward E, Hao Y, Xu J, Murray T, et al. Cancer statistics, 2008. CA Cancer J Clin 2008;58:71–96.
2. Mariani A, Webb MJ, Keeney GL, Aletti G, Podratz KC. Assessment of prognostic factors in stage IIIA endometrial cancer. Gynecol Oncol 2002;86:38–44.
3. Havrilesky LJ, Cragun JM, Calingaert B, Alvarez Secord A, Valea FA, Clarke-Pearson DL, et al. The prognostic significance of positive peritoneal cytology and adnexal/serosal metastasis in stage IIIA endometrial cancer. Gynecol Oncol 2007;104:401–5.
4. Preyer O, Obermair A, Formann E, Schmid W, Perrin LC, Ward BG, et al. The impact of positive peritoneal washings and serosal and adnexal involvement on survival in patients with stage IIIA uterine cancer. Gynecol Oncol 2002;86:269–73.
5. Slomovitz BM, Ramondetta LM, Lee CM, Oh JC, Eifel PJ, Jhingran A, et al. Heterogeneity of stage IIIA endometrial carcinomas: implications for adjuvant therapy. Int J Gynecol Cancer 2005;15:510–6.
6. Saga Y, Imai M, Jobo T, Kuramoto H, Takahashi K, Konno R, et al. Is peritoneal cytology a prognostic factor of endometrial cancer confined to the uterus? Gynecol Oncol 2006;103:277–80.
7. Grimshaw RN, Tupper WC, Fraser RC, Tompkins MG, Jeffrey JF. Prognostic value of peritoneal cytology in endometrial carcinoma. Gynecol Oncol 1990;36:97–100.
8. Ebina Y, Hareyama H, Sakuragh N, Yamamoto R, Furuya M, Sogame M, et al. Peritoneal cytology and its prognostic value in endometrial carcinoma. Int Surg 1997;82:244–8.
9. Tebeu PM, Popowski GY, Verkooijen HM, Casals J, Lüdicke F, Zeciri G, et al. Impact of peritoneal cytology on survival of endometrial cancer patients treated with surgery and radiotherapy. Br J Cancer 2003;89:2023–6.
10. Fadare O, Mariappan MR, Hileeto D, Wang S, McAlpine JN, Rimm DL. Upstaging based solely on positive peritoneal washing does not affect outcome in endometrial cancer. Mod Pathol 2005;18:673–80.
11. Dede M, Yenen MC, Goktolga U, Duru NK, Guden M, Dilek S, et al. Is adjuvant therapy necessary for peritoneal cytology-positive surgical-pathologic Stage I endometrial cancer? Preliminary results. Eur J Gynaecol Oncol 2004;25:591–3.
12. Turner DA, Gershenson DM, Atkinson N, Sneige N, Wharton AT. The prognostic significance of peritoneal cytology for stage I endometrial cancer. Obstet Gynecol 1989;74:775–80.
13. Obermair A, Geramou M, Tripcony L, Nicklin JL, Perrin L, Crandon AJ. Peritoneal cytology: impact on disease-free survival in clinical stage I endometrioid adenocarcinoma of the uterus. Cancer Lett 2001;164:105–10.
14. Lee CM, Slomovitz BM, Greer M, Sharma S, Gregurich MA, Burke T, et al. Practice patterns of SGO members for stage IIIA endometrial cancer. Gynecol Oncol 2005;98:77–83.
15. National Cancer Institute. Surveillance, Epidemiology, and End Results. SEER*Stat Database: Incidence-SEER 9 Regs Limited-Use, Nov 2006 Sub (1973-2004), National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch, released April 2007, based on the November 2006 submission. Available at: http://seer.cancer.gov
. Retrieved October DD, 2007.
16. National Cancer Institute. Surveillance, Epidemiology, and End Results. Overview of the SEER program. Available at: http://seer.cancer.gov/about/
. Retrieved October DD, 2007.
17. Frey CM, McMillen MM, Cowan CD, Horm JW, Kessler LG. Representativeness of the surveillance, epidemiology, and end results program data: recent trends in cancer mortality rates. J Natl Cancer Inst 1992;84:872–7.
19. Tebeu PM, Popowski Y, Verkooijen HM, Bouchardy C, Ludicke F, Usel M, et al. Positive peritoneal cytology in early-stage endometrial cancer does not influence prognosis. Br J Cancer 2004;91:720–4.
20. Takeshima N, Nishida H, Tabata T, Hirai Y, Hasumi K. Positive peritoneal cytology in endometrial cancer: enhancement of other prognostic indicators. Gynecol Oncol 2001;82:470–3.
21. Kasamatsu T, Onda T, Katsumata N, Sawada M, Yamada T, Tsunematsu R, et al. Prognostic significance of positive peritoneal cytology in endometrial carcinoma confined to the uterus. Br J Cancer 2003;88:245–50.
22. Denschlag D, Tan L, Patel S, Kerim-Dikeni A, Souhami L, Gilbert L. Stage III endometrial cancer: preoperative predictability, prognostic factors, and treatment outcome. Am J Obstet Gynecol 2007;196:546.e1–7.