FIGO staging of endometrial cancer: 2023

Many advances in the understanding of the pathologic and molecular features of endometrial cancer have occurred since the FIGO staging was last updated in 2009. Substantially more outcome and biological behavior data are now available regarding the several histological types. Molecular and genetic findings have accelerated since the publication of The Cancer Genome Atlas (TCGA) data and provide improved clarity on the diverse biological nature of this collection of endometrial cancers and their differing prognostic outcomes. The goals of the new staging system are to better define these prognostic groups and create substages that indicate more appropriate surgical, radiation, and systemic therapies.


| INTRODUC TI ON
Since the publication of the last FIGO staging system for endometrial cancer in 2009, a considerable amount of new information has emerged that better defines the pathology and molecular findings as they relate to the type of endometrial carcinoma. In addition, new treatments, results of clinical trials, and prognostic and survival data that correlate with pathologic and surgical findings have been reported. Therefore, the FIGO Committee on Women's Cancer determined that modifications and updates in the staging system were warranted to reflect these new findings and data (Tables 1 and 2).

| Tumor grade
Characterization of histological grade is very important in both the initial biopsy/curettage and the final hysterectomy specimen, especially in EEC and in NSMP. 3  For EECs, grading is prognostically significant. 6,7 The grading criteria for EECs are primarily based on architectural features. 8 In this revised FIGO staging scheme, the binary approach of WHO Classification of Tumors, Female Genital Tumors 1 has been adopted.
In brief, low-grade EECs are subdivided into grade 1 and 2 tumors, which exhibit up to 5% and 6%-50% solid non-glandular growth, respectively. 8 In contrast, high-grade EECs (grade 3) are characterized by 50% or more solid component. This binary grading system allows easier clinical decision-making and has improved reproducibility, 9 but it is prudent to remember that the three-tiered system is still

| Myometrial invasion
The extent of myometrial invasion has long been recognized as an essential prognostic risk factor. 10 It is recommended that the assessment of the percentage of myometrium involved should be expressed as the percentage of the overall myometrial thickness infiltrated by carcinoma using three categories: none; <50%; or ≥50%. [11][12][13][14] The assessment of tumor invasion from adenomyosis is a controversial issue without strong scientific evidence. 15

| Lymphovascular space invasion (LVSI)
LVSI should be assessed at the invasive front of the tumor. 15,16 It is crucial to distinguish LVSI from mimickers, such as a microcystic elongated and fragmented (MELF) pattern of myometrial invasion and retraction artifacts [17][18][19] that may occur in the setting of minimally invasive surgery. It is very important to distinguish "substantial" or "extensive" LVSI from "focal" or "no" LVSI. [20][21][22] Although determination of the precise number of involved vessels to discriminate between focal and extensive/substantial requires additional scientific evidence, for staging purposes the recommendation by WHO 2020 (≥5 vessels) is adopted. 1 • Serous adenocarcinomas, clear cell adenocarcinomas, mesonephric-like carcinomas, gastrointestinal-type mucinous endometrial carcinoma, undifferentiated carcinomas, and carcinosarcomas are considered high-grade by definition. For EECs, grade is based on the proportion of solid areas: low grade = grade 1 (≤5%) and grade 2 (6%-50%); and high grade = grade 3 (>50%). Nuclear atypia excessive for the grade raises the grade of a grade 1 or 2 tumor by one. The presence of unusual nuclear atypia in an architecturally low-grade tumor should prompt the evaluation of p53 and consideration of serous carcinoma. Adenocarcinomas with squamous differentiation are graded according to the microscopic features of the glandular component.
• It should be noted that high-grade EECs (grade 3) are a prognostically, clinically, and molecularly heterogenous disease, and the tumor type that benefits most from applying molecular classification for improved prognostication and for treatment decision-making. 3 Without molecular classification, high-grade EECs cannot appropriately be allocated to a risk group and thus molecular profiling is particularly recommended in these patients. For practical purposes and to avoid undertreatment of patients, if the molecular classification is unknown, high-grade EECs were grouped together with the aggressive histological types in the actual FIGO classification. f Micrometastases are considered to be metastatic involvement (pN1 (mi)). The prognostic significance of isolated tumor cells (ITCs) is unclear. The presence of ITCs should be documented and is regarded as pN0(i+). According to TNM8, macrometastases are >2 mm in size, micrometastases are 0.2-2 mm and/or >200 cells, and isolated tumor cells are ≥0.2 mm and ≤200 cells. 33  The recognition of myometrial invasion and identification of LVSI in tumor tissue is dependent on appropriate sampling. Therefore, it is important to consider ISGYP recommendations, which state that one section per centimeter of the largest tumor dimension will suffice. 23

| Cervical stromal invasion
Cervical stromal invasion is subjected to significant inter-observer variation 24,25 and strict criteria are recommended. Any invasion of the cervical stroma, identified at the level of or deeper than a benign endocervical crypt, should be considered cervical stromal invasion.
Cervical glandular extension is not considered for staging.

| Adnexal involvement
Adnexal involvement has an impact on overall survival. 26,27 In the past, it was considered necessary to distinguish between endometrial carcinoma with ovarian metastasis and synchronous primary tumors of the endometrium and the ovary. In the case of highgrade tumors, ovarian involvement is almost always categorized as metastatic. However, for low-grade EECs, the situation is complex.
Recent molecular studies have shown that there is a clonal relationship between the endometrial and ovarian tumor in the vast majority of cases, suggesting that the tumor arises in the endometrium, and secondarily extends to the ovary. 28

Stage designation Molecular findings in patients with early endometrial cancer (Stages I and II after surgical staging)
Stage IAm POLEmut POLEmut endometrial carcinoma, confined to the uterine corpus or with cervical extension, regardless of the degree of LVSI or histological type Stage IICm p53abn p53abn endometrial carcinoma confined to the uterine corpus with any myometrial invasion, with or without cervical invasion, and regardless of the degree of LVSI or histological type Abbreviation: LVSI, lymphovascular space involvement. a When feasible, the addition of molecular subtype to the staging criteria allows a better prediction of prognosis in a staging/prognosis scheme. The performance of complete molecular classification (POLEmut, MMRd, NSMP, p53abn) is encouraged in all cases of endometrial cancer for prognostic risk-group stratification and as potential influencing factors of adjuvant or systemic treatment decisions. Molecular subtype assignment can be done on a biopsy, in which case it need not be repeated on the hysterectomy specimen. When performed, these molecular classifications should be recorded in all stages.
• Good prognosis: pathogenic POLE mutation (POLEmut) • Intermediate prognosis: mismatch repair deficiency (MMRd)/microsatellite instability and no specific molecular profile (NSMP) • Poor prognosis: p53 abnormal (p53abn)When the molecular classification is known: • FIGO Stages I and II are based on surgical/anatomical and histological findings. In case the molecular classification reveals POLEmut or p53abn status, the FIGO stage is modified in the early stage of the disease. This is depicted in the FIGO stage by the addition of "m" for molecular classification, and a subscript is added to denote POLEmut or p53abn status, as shown below.

| Uterine serosal involvement
By following ISGYP recommendations, 15 uterine serosal involvement is defined as a tumor reaching submesothelial fibroconnective tissue or the mesothelial layer, regardless of whether tumor cells may or may not be present on the serosal surface of the uterus.

| Lymph node status
Lymph node status is an important prognostic factor for endometrial carcinoma. According to TNM8, 33  Several studies have demonstrated the prognostic value of this TCGA-surrogate approach. POLEmut denotes a favorable prognosis.

MMRd and NSMP indicate an intermediate prognosis, while p53abn
indicates a poor prognosis. Most notably, data suggest that carcinomas falling into the POLEmut group may benefit from de-escalation of postoperative adjuvant therapy because of the consistently better outcome in these cases. In contrast, p53abn has a much worse prognosis, suggesting that some form of increased intensive therapy may be of benefit. Improved risk assessment by integrating molecular and clinicopathological factors in endometrial carcinoma has been demonstrated by many studies. [42][43][44][45][46][47][48][49][50] Furthermore, the TCGA surrogate approach has been verified by the molecular portion of PORTEC 3. 38 There is a small subset of tumors (approximately 5%) that combine more than one molecular feature (e.g. POLEmut and p53abn or MMRd and p53abn), and they are referred to as "multiple clas-  There is a subset of patients with low-grade endometrioid carcinomas involving the endometrium and the ovaries, which are associated with a good prognosis. [63][64][65] They were previously described as synchronous independent tumors, but molecular analysis has established a common clonal origin. 28 The absence of LVSI and focal LVSI have been related to a good prognosis in opposition to substantial/extensive LVSI in low-grade EECs restricted to the uterus. 22,[60][61][62] The criteria for LVSI follow the rules of WHO. 1 Accordingly, LVSI should fall into one of the following three categories: "LVSI negative" (0 vessels); "LVSI focal" (<5 vessels); or "LVSI substantial/extensive" (≥5 vessels).

| Stage II
The revised staging system includes major changes to Stage II.
The number of women with Stage II tumors will markedly increase under the new staging system. Stage IIA tumors include non-aggressive histotlogical that have invasion of the cervical stroma.
Stage IIB now represents cases that include non-aggressive histological types with substantial LVSI as defined by the WHO 2021 report, regardless of local tumor spread. An extensive body of literature supports these findings. Randomized trials, prospective cohort studies, large database series, and single-institution reports consistently demonstrate that LVSI is an independent and strong prognostic factor for the recurrence of endometrial carcinoma. [66][67][68][69] A retrospective registry study of more than 1500 patients from

| Stage III
In Stage III, the tumor has spread locally or regionally. The revised subclassifications aim to better reflect the clinical picture and prognosis and enable a more appropriate treatment decision-making process. The differences from the previous staging system are sum- Third, Stage IIIC is further divided into micrometastasis (IIIC1i, IIIC2i) and macrometastasis to the lymph nodes (IIIC1ii, IIIC2ii), while isolated tumors cells (ITC) are not considered metastatic and regarded as pN0(i+). The substaging is based on the better prognosis in those patients who have micrometastasis to the lymph nodes. [73][74][75][76][77] This subcategorization also reflects the increasing utilization of the sentinel lymph node technique and ultrastaging, which allows im-

| Stage IV
The main change to this part of the

| DISCUSS ION
The purpose of this revision of the FIGO endometrial cancer staging system is to incorporate the essential new published evidence as summarized above. The goal is to improve the clarity of the diverse biological nature of endometrial carcinomas with differing prognostic outcomes, better define these prognostic groups, and create substages that yield more appropriate surgical, radiation, and systemic therapies. As with all staging systems, the evolution of the updated classification must be based on the results of clinical studies.
The committee felt that the risk stratification, including the molecular classification that has recently been developed by ESGO, ESTRO, and ESP helps to better define the prognosis and therapeutic approaches for these diseases. 80,81 Therefore, it was determined that notation of the molecular classification, when performed, should be included to stage the patient's disease.
Notwithstanding the fact that these tests may not be available in some settings, the molecular findings are sufficiently prognostic that treatment might be modified in those patients for whom this information is obtained.
The inclusion of molecular measures for endometrial cancer follows the work done with breast cancer staging in 2018, when, along with tumor grade, several molecular assays results-estrogen receptor status, progesterone receptor status, and Her2neu-were added to the staging system to reflect the impact on prognosis of these significant molecular parameters. 34 In summary, the current modifications to the endometrial staging system have been made to further define the differences in progno- Mutch, and Nicole Concin were responsible for the conceptualization, investigation, methodology, writing the original draft, and reviewing and editing the manuscript.

ACK N OWLED G M ENTS
We dedicate this staging system to all women diagnosed with endometrial cancer in the hope that it will help to improve their care and treatment. The authors and committee members express their gratitude to Jeanne A. Conry, MD, PhD, FIGO President, and Mary Ann Lumsden, MD, FIGO CEO, for their guidance and support.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analyzed in this study.