Endometrial cancer is the most common gynecologic malignancy in the United States, with the disease diagnosed in an estimated 47,000 women in 2012. Because most women had their diagnosis at an early stage, only approximately 8000 women died of their disease in 2012.1 The International Federation of Gynecology and Obstetrics (FIGO) endometrial cancer staging system was changed from a clinical to a surgical staging system in 1988 based on large clinicopathologic studies highlighting the importance of pathologic findings.2,3 The 1988 FIGO system for stage I endometrial cancer had 3 subgroups: IA, no myometrial invasion; IB, less than half myometrial invasion; and IC, half or greater myometrial invasion. The revised, 2009 FIGO system combined stages IA and IB, resulting in 2 subgroups: IA, no or less than half myometrial invasion; and IB, half or greater myometrial invasion.4
Most patients with endometrial cancer have stage I disease at diagnosis, but this is not a homogeneous population. Many of these women will be cured of their disease, but some will recur and may die of their disease. Both the 1988 and 2009 FIGO staging systems for stage I endometrial cancer are based solely on myometrial invasion. A recent publication by The Cancer Genome Atlas Research Network on endometrial carcinoma highlighted the importance of tumortyping in overall survival prognosis.5 The authors found 4 endometrial cancer types: DNA polymerase epsilon (POLE) ultramutated, microsatellite instability hypermutated, copy number low, and copy number high. The copy number high group (serous-like), which was composed of mostly serous tumors and grade 3 endometrioid tumors, had the worst prognosis. This prognostic histologic information can be incorporated into a revisedstaging system. In addition, previous FIGO staging systems for stage I endometrial cancer do not take into account whether or not lymph nodes were removed, which continues to be a controversial area in endometrial cancer surgical staging. Moreover, the grading system remains unchanged, with 3 grade categories.
Accommodating for the distinctive characteristics of different disease types (histological subtypes) is not novel. International Federation of Gynecology and Obstetrics 2009 recognizes the distinctive characteristics of gynecologic histologic subtypes, providing different staging schemes for uterine carcinoma and carcinosarcoma than for uterine leiomyosarcoma, endometrial stromal sarcoma, and Mullerian adenosarcoma. Our data show that incorporating histological subtype and whether or not lymph nodes were removed (“nodes known” vs “nodes unknown”) improves prognostic accuracy for patients with FIGO stage I endometrial carcinoma.
The purpose of this study was to develop a new staging system for stage I endometrial cancer that, in addition to myometrial invasion, takes into consideration disease type (ie, histological subtype), grade (using a binary system instead of a 3-grade system), and whether or not lymph nodes were examined. Such a staging system for stage I endometrial cancers would reflect the current reality of surgical staging of this disease, given the controversy over the role of lymphadenectomy in endometrial cancer staging. We aimed to compare the performance of the new system to the 1988 and 2009 FIGO staging systems.
MATERIALS AND METHODS
With the approval of the institutional review board of Memorial Sloan-Kettering Cancer Center, all patients who underwent surgery and had a diagnosis of 1988 FIGO stage I endometrial cancer from January 1993 through August 2011 were analyzed. Sarcomas and undifferentiated carcinomas were excluded. Twelve patients had mixed tumor types, and 14 patients had mucinous histology. After review by our expert gynecologic pathologists, it was decided that these would best be analyzed as part of the endometrioid histologic subtype. Standard demographic, clinical, and pathologic data were extracted. We defined endometrioid grade 1 or 2 cancers as “low grade”, and we defined endometrioid grade 3 or nonendometrioid carcinomas (serous, clear cell, and carcinosarcoma) as “high grade”. We proposed the following substaging definitions for stage I endometrial cancers:
- IA. Low-grade carcinoma with less than half myometrial invasion
- IA1: Negative nodes
- IA2: No nodes removed
- IB. High-grade carcinoma with no myometrial invasion
- IB1: Negative nodes
- IB2: No nodes removed
- IC. Low-grade carcinoma with half or greater myometrial invasion
- IC1: Negative nodes
- IC2: No nodes removed
- ID. High-grade carcinoma with any myometrial invasion
- ID1: Negative nodes
- ID2: No nodes removed
The vital status of each patient was recorded. Overall survival (OS) was calculated from the date of surgery to either date of last follow-up or death. The Kaplan-Meier method was used to estimate OS probabilities. Univariate OS analysis was performed using the log-rank test, and the Cox proportional hazards model was fitted univariately to estimate the hazard ratio for each covariate separately.
We compared the 1988, 2009, and our proposed endometrial cancer stage I system using the concordance probability, which is a measure of predictive accuracy.6 Analogous to area under the receiver operating curve, concordance probability estimate (CPE) can range from perfect concordance at 1.0 to perfect discordance at 0.0. More specifically, a CPE of 50% implies no predictive accuracy because for 2 randomly selected patients, there is a 50% chance that the patient with the higher predicted probability by the staging system will have longer survival than the other patient, which is no better than a coin flip. To prevent against overfitting the model, the bootstrap-corrected CPE was also reported.7
A total of 1843 women met study criteria and were analyzed. Demographic information is summarized in Table 1. More than 80% of the patients were alive without evidence of disease at last follow-up. Among the 1592 survivors, the median follow-up time was 49.7 months. The median age was 61 years, and the median body mass index was 29.9 kg/m2. Eighty-seven percent of the women had endometrioid carcinoma; and the remainder had serous carcinoma, carcinosarcoma (malignant mixed Mullerian tumors), or clear cell carcinoma. During the study period, 67.8% of the patients had lymph nodes removed. Among the 1249 patients with lymph nodes excised, the median lymph node count was 16. The median number of pelvic and para-aortic lymph nodes removed was 13 and 3, respectively.
On univariate analysis, age, grade, disease type (ie, histological subtype), and staging using the 1988, the 2009, and the new proposed staging systems were all significantly associated with OS (Table 2). The removal of lymph nodes was not statistically significant (P = 0.103). The 5-year OS rates for the 1988 staging system were the following: IA, 95.4%; IB, 88.4%; and IC, 81.1% (P < 0.001). The Kaplan-Meier OS curves stratified by the 1988 FIGO stage I system are shown in Figure 1A. The 5-year OS rates for the 2009 staging system were IA, 91.5%; and IB, 81.1% (P < 0.001), as illustrated in Figure 1B. When patients were restaged with our proposed system for stage I disease, the 5-year OS rates significantly differed (P < 0.001) as follows: IA1, 96.7%; IA2, 92.2%; IB1, 92.2%; IB2, 76.4%; IC1, 83.9%; IC2, 78.6%; ID1, 81.1%; and ID2, 68.8%. The Kaplan-Meier analysis for the proposed new system is presented in Figure 2.
A cross-tabulation of the 1988, the 2009, and the proposed endometrial cancer stage I systems is shown in Table 3. Most patients had no or less than 50% myometrial invasion and had grade 1 or 2 endometrioid adenocarcinoma. Thirteen percent (239/1843) of the patients who would otherwise have been classified as stage IA using the 2009 system were classified as stage ID using the proposed system.
The bootstrap-corrected CPE for the proposed staging system was 0.627 (95% confidence interval [CI], 0.590–0.664) and was superior to the CPE of 0.530 (95% CI, 0.516–0.544) for the 2009 FIGO staging system. The bootstrap-corrected CPE for the new system was higher than the CPE of 0.615 (95% CI, 0.582–0.648) for the 1988 system, but their confidence intervals overlapped (Table 4).
The Proposed Staging System Is More Prognostically Powerful Than Existing and Previous Systems
Gynecologic cancer stage definitions are designed to provide relevant clinical information, inform prognosis, and assist in guiding patient management decisions.8 It has been stated that a good staging system must have 3 basic characteristics: it must be valid, reliable, and practical.9 It should group patients who experience similar outcomes together, ensure that identical cases are always assigned to the same category, and should reflect clinical practice. Because endometrial cancer often presents with postmenopausal or irregular vaginal bleeding, most women are diagnosed at an early stage. However, patients with early-stage endometrial cancer comprise a heterogeneous population with regard to histologic subtype, tumor grade, and prognosis. A stage I definition that divides this group into 2 groups based solely on myometrial invasion, as with the 2009 FIGO stage I system, is inadequate for providing accurate prognostic information. Additional relevant clinical information should be incorporated to place patients in the appropriate prognostic group and manage them appropriately. We propose a staging system for stage I endometrial cancer that goes beyond myometrial invasion and incorporates disease type (ie, histological subtype), grade (a binary system of low versus high grade), and whether or not lymph nodes were evaluated.
The FIGO staging system for endometrial cancer has undergone several revisions through the years based on emerging information, but these changes have not always improved the performance of the staging system. Specific to stage I disease, our group questioned whether the 1988 FIGO stages IA and IB should have been altered.10 Using data from more than 1600 women, the OS of women with stage I endometrial cancer was analyzed to evaluate the performance of the 1988 and 2009 FIGO staging systems for stage I endometrial cancer. Both systems provided statistically significant survival differences between substages, but the adjusted concordance probability for the 1988 stage I group was 0.612 compared to 0.536 for the 2009 stage I group. The authors concluded that the revised 2009 system eliminated the most favorable group with no myometrial invasion and did not improve its predictive ability over the 1988 system, emphasizing the importance of individualized risk prediction models.
The landmark article by Creasman et al. describing the surgical pathologic spread patterns of endometrial cancer confirmed previous findings2,11 and laid the groundwork for the change in endometrial cancer staging from a clinical to a surgical staging system; they reported that the depth of myometrial invasion was an important predictor of extrauterine disease.3 Myometrial invasion was divided into 4 groups: endometrium only, inner third, middle third, and deep third. Invasion was significantly associated with frequency of nodal metastasis. The 1988 and 2009 FIGO stage I systems are based solely on depth of myometrial invasion, which is an overly simplified division of patients with stage I endometrial cancer.
Grade was another significant risk factor for nodal metastasis in the article by Creasman et al and continues to commonly be referred to in combination with depth of myometrial invasion to describe the frequency of pelvic and para-aortic node metastases.3 For instance, no patients with tumor limited to the endometrium and grade 1 histology had positive pelvic or para-aortic nodes, whereas patients with deep myometrial invasion and grade 3 histology had a 34% risk of pelvic node metastasis and a 23% risk of para-aortic node metastasis. Clearly, grade and depth of invasion are critical pieces of information in determining prognosis and management, and our proposed stage I endometrial cancer system takes both grade and myometrial invasion into consideration. It further simplifies the grading process into a binary system (low vs high grade).
Endometrial Carcinomas Comprise Different Disease Types
In 1983, Bokhman12 described 2 pathogenetic types of endometrial carcinoma. Type 1 endometrial cancer was driven by hyperestrinism and often presented in obese women with hyperlipidemia, whereas type 2 endometrial cancer was not associated with unopposed estrogen. Moreover, type 2 endometrial cancer was noted to be much more aggressive, with 66% having deep myometrial invasion and 63% being poorly differentiated, compared to type 1 in which 69% had superficial invasion and 82% were grade 1 or 2. This translated into a lower 5-year OS in type 2 (59%) compared to type 1 (86%) endometrial cancer. There is now a wealth of reported data that support the idea that endometrial carcinoma is composed of at least 2 distinctive disease types, each of which is characterized by different epidemiologic risk factors, morphology, immunophenotype, genotype, dissemination patterns, recurrence-free and overall survivals, and, perhaps, therapeutic response.5,12–15
Disease Type Should Be Accounted for in an Endometrial Cancer Staging System
Authors have argued that the 2009 FIGO staging system for use in uterine serous carcinoma is oversimplified.16 Seward et al17 evaluated the prognostic ability of the revised 2009 staging system compared to the 1988 system in uterine serous carcinoma and concluded that the 2009 criteria did not adequately delineate survival for uterine serous carcinoma in early-stage disease. The patients with no myometrial invasion had the most favorable prognosis, which the authors argued warranted a separate substage. Distinguishing serous and low-grade endometrioid carcinomas without myometrial invasion in a staging scheme allows for the discrimination of patients with favorable and potentially highly unfavorable clinical outcomes. Unlike low-grade endometrioid carcinomas without myometrial invasion, serous carcinomas lacking myometrial invasion are frequently metastatic at presentation, and suboptimal staging strategies might fail to recognize patients at highest risk of distant, particularly peritoneal, failure. Once low stage is verified after a comprehensive staging surgery, our proposal allows for the separation of serous carcinomas into 2 groups stratified by the presence of myometrial invasion, thereby preserving the best prognostic group.
Reports that compare clinical outcomes in FIGO grade 3 endometrioid and serous carcinomas do not uniformly conclude that these tumor types are comparable, but there are several noteworthy studies that describe similar clinical outcomes. During a study period using contemporary therapeutic interventions for high-risk endometrial carcinoma, Ayeni et al18 controlled for disease stage and reported no difference in OS when comparing grade 3 endometrioid, serous, and clear cell carcinomas. Another recent publication by Voss et al similarly found that disease-specific and recurrence-free survivals were similar between grade 3 endometrioid endometrial cancer and serous cancer or clear cell cancer. Previous reports have also demonstrated similar clinical outcomes when comparing patients with grade 3 endometrioid endometrial carcinoma to women with uterine serous or clear cell histology.19,20 Our proposed stage I endometrial cancer system recognizes the similar clinical behavior of grade 3 endometrioid endometrial cancer and the type 2 cancers and therefore combines grade 3 endometrioid carcinoma with nonendometrioid carcinoma, which in this study included serous, clear cell, and malignant mixed Mullerian tumors.
Furthermore, recent data from The Cancer Genome Atlas Research Network’s study of endometrial carcinoma indicate significant genotypic overlap between FIGO grade 3 endometrioid and serous carcinomas, using historical criteria for diagnosis.5 This work has also provided outcomes data that emphasize the distinctive and aggressive nature of serous carcinomas and related tumors that mapped to the copy number high cluster. This cluster included 94% of serous tumors, 62% of mixed-histology tumors, and 24% of grade 3 endometrioid tumors. Our staging system provisionally combines grade 3 endometrioid carcinoma with nonendometrioid carcinoma while we await the development of precise, clinically validated histological criteria that separate the most clinically aggressive tumor types (serous carcinoma, carcinosarcoma, and serous-like endometrioid carcinoma) from otherwise intermediate-risk carcinomas, such as prototypical FIGO grade 3 endometrioid carcinoma.
Thoroughness of Lymph Node Evaluation Is Prognostically Predictive
An insightful commentary by Creasman21 described the current controversy of the role of lymphadenectomy in the management of endometrial cancer and posed 2 questions: do all patients require lymphadenectomy, and who if anyone needs adjuvant therapy? After 2 randomized controlled trials failed to demonstrate improved survival after pelvic lymphadenectomy, the importance of lymphadenectomy began to be considered more diagnostic than therapeutic.22,23 Lymphadenectomy is critical for accurate staging and adjuvant treatment decisions. The diagnostic importance of lymph node evaluation supports the increasing use of sentinel lymph node mapping in endometrial cancer and the implementation of a sentinel lymph node algorithm to excise the nodes that are most likely to represent the lymph node status of each patient.24,25 However, real-life clinical practice represents a spectrum of patterns, ranging from no lymphadenectomy to a full pelvic and para-aortic lymph node dissection. A patient classified as stage I with pathologically negative lymph nodes is different from a woman assigned to stage I who never had her lymph nodes assessed, and this distinction must be reflected in the staging system. Separating staging subgroups by the evaluation of lymph nodes is not novel but is informative. The American Joint Committee on Cancer TNM staging for uterine cancer has an Nx category to designate that no lymph nodes were assessed. Our proposed stage I endometrial cancer staging system includes subgroups that take into account whether or not lymph nodes were excised for pathologic evaluation and were negative. We do not require a specific number of nodes owing to the tremendous variation in practices and the difficulty in reaching consensus on what is an adequate lymphadenectomy for staging.26,27
By incorporating the well-recognized and important variables of disease type, grade, myometrial invasion, and whether or not lymph nodes were removed (a measure of surgical staging), our proposed staging system for stage I endometrial cancer has a superior predictive ability over the 2009 FIGO staging system and provides additional relevant clinical information over both the 1988 and 2009 systems. In the era of personalized medicine, it logically follows that we should use more of the information available to us to guide and individualize care for the heterogeneous population of women who have early-stage endometrial cancer. Adopting this new staging system would provide more accurate prognostic information and may assist in treatment planning. The new system also more accurately reflects the reality of surgical staging in which patients may not have any nodes pathologically assessed. The proposed staging system is not meant to be comprehensive or definitive but rather emphasizes the continuous need to revise the existing staging system as more is discovered about endometrial carcinoma. The proposed groupings are preliminary and could be reformulated for ease of use. Further work is needed to externally validate the proposed staging system.
1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin. 2012; 62: 10–29.
2. Boronow RC, Morrow CP, Creasman WT, et al. Surgical staging in endometrial cancer
: clinical-pathologic findings of a prospective study. Obstet Gynecol. 1984; 63: 825–832.
3. Creasman WT, Morrow CP, Bundy BN, et al. Surgical pathologic spread patterns of endometrial cancer
. A Gynecologic Oncology Group Study. Cancer. 1987; 60 (suppl 8): 2035–2041.
4. Pecorelli S. Revised FIGO
staging for carcinoma of the vulva, cervix, and endometrium. Int J Gynaecol Obstet. 2009; 105: 103–104.
5. Cancer Genome Atlas Research Network, Kandoth C, Schultz N, et al. Integrated genomic characterization of endometrial carcinoma. Nature. 2013; 497: 67–73.
6. Gonen M, Heller G. Concordance probability and discriminatory power in proportional hazards regression. Biometrika. 2005; 92: 965–970.
7. Harrell FE Jr, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 1996; 15: 361–387.
8. Benedet JL, Pecorelli S. Why cancer staging? FIGO
26th annual report on the results of treatment in gynecological cancer. Int J Gynaecol Obstet. 2006; 95 (suppl 1): S3.
9. Odicino F, Pecorelli S, Zigliani L, et al. History of the FIGO
cancer staging system. Int J Gynecol Obstet. 2009; 101: 205–210.
10. Abu-Rustum NR, Zhou Q, Iasonos A, et al. The revised 2009 FIGO
staging system for endometrial cancer
: should the 1988 FIGO
stages IA and IB be altered? Int J Gynecol Cancer. 2011; 21: 511–516.
11. DiSaia PJ, Creasman WT, Boronow RC, et al. Risk factors and recurrent patterns in stage I
endometrial carcinoma. Am J Obstet Gynecol. 1984; 151: 1009–1015.
12. Bokhman JV. Two pathogenetic types of endometrial carcinoma. Gynecol Oncol. 1983; 15: 10–17.
13. Mhawech-Fauceglia P, Yan L, Liu S, et al. ER /PR /TFF3 /IMP3 immunoprofile distinguishes endometrioid from serous and clear cell carcinomas of the endometrium: a study of 401 cases. Histopathology. 2013; 62: 976–985.
14. Paik D, Cocco E, Bellone S, et al. Higher sensitivity to patupilone versus paclitaxel chemotherapy in primary uterine serous papillary carcinoma cell lines with high versus low HER-2/neu expression in vitro. Gynecol Oncol. 2010; 119: 140–145.
15. Kelly MG, O’Malley DM, Hui P, et al. Improved survival in surgical stage I
patients with uterine papillary serous carcinoma (UPSC) treated with adjuvant platinum-based chemotherapy. Gynecol Oncol. 2005; 98: 353–359.
16. Winer I, Mahdi H, Bandyopadhyay S, et al. Correlation of tumor size with other prognostic factors in uterine serous carcinoma: a large multi-institutional study. Gynecol Oncol. 2013; 128: 316–321.
17. Seward S, Ali-Fehmi R, Munkarah AR, et al. Outcomes of patients with uterine serous carcinoma using the revised FIGO
staging system. Int J Gynecol Cancer. 2012; 22: 452–456.
18. Ayeni TA, Bakkum-Gamez JN, Mariani A, et al. Comparative outcomes assessment of uterine grade 3 endometrioid, serous, and clear cell carcinomas. Gynecol Oncol. 2013; 128: 71–76.
19. Soslow RA, Bissonnette JP, Wilton A, et al. Clinicopathologic analysis of 187 high-grade endometrial carcinomas of different histologic subtypes: similar outcomes belie distinctive biologic differences. Am J Surg Pathol. 2007; 31: 979–987.
20. Alektiar KM, McKee A, Lin O, et al. Is there a difference in outcome between stage I
-II endometrial cancer
of papillary serous/clear cell and endometrioid FIGO
Grade 3 cancer? Int J Radiat Oncol Biol Phys. 2002; 54: 79–85.
21. Creasman WT. The current status of lymphadenectomy in the management of endometrial cancer
. Womens Health (Lond Engl). 2011; 7: 33–35.
22. Benedetti Panici P, Basile S, Maneschi F, et al. Systematic pelvic lymphadenectomy vs. no lymphadenectomy in early-stage endometrial carcinoma: randomized clinical trial. J Natl Cancer Inst. 2008; 100: 1707–1716.
23. ASTEC Study Group, Kitchener H, Swart AM, et al. Efficacy of systematic pelvic lymphadenectomy in endometrial cancer
(MRC ASTEC trial): a randomised study. Lancet. 2009; 373: 125–136.
24. Barlin JN, Khoury-Collado F, Kim CH, et al. The importance of applying a sentinel lymph node mapping algorithm in endometrial cancer
staging: beyond removal of blue nodes. Gynecol Oncol. 2012; 125: 531–535.
25. Khoury-Collado F, Murray MP, Hensley ML, et al. Sentinel lymph node mapping for endometrial cancer
improves the detection of metastatic disease to regional lymph nodes. Gynecol Oncol. 2011; 122: 251–254.
26. Cormier B, Sauthier P, Lussier C, et al. Determinants of lymph node count in endometrial cancer
surgical staging. Int J Gynecol Cancer. 2012; 22: 1361–1366.
27. Abu-Rustum NR, Iasonos A, Zhou Q, et al. Is there a therapeutic impact to regional lymphadenectomy in the surgical treatment of endometrial carcinoma? Am J Obstet Gynecol. 2008; 198: 457.e1–457.e5; discussion 457.e5-457.e6.