Endometrial carcinoma is the most frequent cancer of the female genital tract in the Western world [1]. In 1983, Bokhman described two different clinical types of endometrial carcinoma [2]: type 1 endometrial carcinoma (65%) occurred in obese premenopausal or perimenopausal women with unopposed estrogen, hyperlipidemia, and endometrial hyperplasia, whereas women with type 2 endometrial carcinoma (35%) were older and had adjacent atrophic endometrium or atrophic endometrial polyps. Type 1 endometrial carcinomas were well or moderately differentiated tumors (grades 1 and 2) that exhibited only superficial myometrial invasion and had favorable prognosis (86% 5-year survival rate); in contrast, type 2 endometrial carcinomas, which included uterine papillary serous carcinomas previously described by Hendrickson et al. [3], were poorly differentiated cancers (grade 3) that often invaded the myometrium and were associated with adverse prognosis even in the absence of deep myoinvasion (59% 5-year survival rate). Time has shown that Bokhman's subdivision is essentially conceptual and is not useful for the routine diagnosis of endometrial carcinoma. Subsequently, refined histopathologic features and molecular genetic alterations were added: type 1 (currently estimated in 80%) are endometrioid carcinomas with mutations of PTEN, KRAS, CTNNB1, PIK3CA and microsatellite instability, whereas type 2 (about 20%) includes not only serous carcinomas but also clear cell carcinomas and carcinosarcomas; however, the vast majority of type 2 tumors are serous carcinomas with TP53 mutations, inactivation of p16 and E-cadherin, c-erbB2 amplification, and loss of heterozygosity on several chromosomes, reflecting chromosomal instability [4,5]. Clear cell carcinomas are extremely rare and, in fact, many so-called “clear cell carcinomas” turned out to be serous or endometrioid carcinomas with clear cells, raising the question whether clear cell carcinomas really occur in the corpus. Carcinosarcomas are aggressive tumors that share some clinical features with carcinomas (for instance, lymph nodal spread) but carry a worse prognosis, similar to that of uterine sarcomas [6]; accordingly, in the latest WHO classification, carcinosarcomas are included in the section of endometrial carcinomas as an aggressive subtype which always carries a TP53 mutation [6]. Although the majority of endometrial carcinomas are assigned to type 1 or type 2 subgroups, some cases are difficult to classify; i.e., there are carcinomas with mixed or ambiguous morphology with molecular features of both endometrioid and serous carcinomas (Fig. 1). The 2020 WHO classification of the endometrial epithelial tumors is shown in Fig. 2. In order to refine the histopathologic classification, The Cancer Genome Atlas (TCGA) Research Network [7], based on nucleotide substitution frequencies and patterns, MSI status, and somatic copy number status, described 4 molecular subgroups of endometrial carcinomas with different prognosis (Fig. 3): 1) endometrioid carcinomas with POLE mutations (ultramutated); 2) endometrioid carcinomas with microsatellite instability (hypermutated); 3) endometrioid carcinomas with a low degree of copy-number alterations; and 4) predominantly serous but also serous-like endometrioid carcinomas with a high degree of copy-number alterations and frequent TP53 mutations. Accordingly, an algorithm based on a limited panel of immunohistochemistry (MMR proteins and p53) and POLE mutation analysis was subsequently developed [6,8] (Fig. 4). After 10 years of comprehensive analysis, it appears that the only real contribution of TCGA to the clinical management of these patients would be limited to the uncommon high-grade, low-stage endometrioid carcinomas with POLE exonuclease domain mutations, as these patients could benefit from a de-escalating treatment; knowledge about the other three molecular groups has not changed significantly.

Recently, TCGA molecular subgroups have been integrated into the 2023 International Federation of Gynecology and Obstetrics (FIGO) staging classification which incorporates other non-anatomic parameters like histotype, tumor grade, and lymphovascular space invasion (LVSI) [9]. The new FIGO staging system which is totally different from the previous one, is a complicated and non-intuitive classification that makes its clinical application difficult and does not facilitate correspondence with the 2009 FIGO classification scheme [10]. Contrariwise, a 2019 proposal for updating the staging of endometrial cancer that considered TCGA molecular data complementary to the traditional FIGO anatomical parameters has been overlooked [11]. In this review, we highlight the different controversial points of the new FIGO staging system.