Endometrial cancer: Pathology and classification

INTRODUCTION — Endometrial cancers (ECs) comprise a variety of neoplasms with variable patient outcomes/prognosis.

Traditionally, EC has been classified by histomorphologic features and stratified into the more common, lower-risk, estrogen-driven cancers (type I) and the less common, more clinically aggressive/unfavorable, nonestrogen-driven cancers (type II). However, this approach does not adequately capture the complexity of these neoplasms, and histomorphologic classification has been demonstrated to be poorly reproducible in ECs, particularly for grade and histotype assignment. Thus, the World Health Organization (WHO) and International Federation of Gynecology and Obstetrics (FIGO) have subsequently developed other classification systems that incorporate molecular features to provide more objective, reproducible categorization, and that help inform prognosis and may be used to guide treatment decisions.

The pathology and classification of EC are discussed here; other related content is reviewed separately:

●(See "Overview of resectable endometrial carcinoma".)

●(See "Endometrial carcinoma: Epidemiology, risk factors, and prevention".)

●(See "Endometrial carcinoma: Clinical features, diagnosis, prognosis, and screening".)

●(See "Endometrial carcinoma: Staging and surgical treatment".)

CLASSIFICATION — The classification of EC has evolved over time.

Traditional classification system — In 1983, two clinicopathogenetic types of EC (type I and type II) were described based on endocrine and metabolic influences, pathologic features, and prognosis [1]. In this classification system, type I ECs (the majority of ECs) include those that are estrogen driven, mostly endometrioid histology, lower grade, have less myometrial invasion, and a more favorable prognosis. By contrast, type II ECs comprise a diverse mix of high-grade, clinically aggressive histologies (eg, serous, clear cell), with a poor response rate to progestogens and poor outcomes.

This classification system, although helpful in providing a conceptual framework for understanding EC, does not capture the biological diversity or diversity of clinical outcomes of some histotypes (eg, grade 3 endometrioid cancers, mixed ECs), which do not always fit neatly into one of the two categories (figure 1).

WHO categorization — The 5^th edition of the World Health Organization (WHO) system, published in 2020, histologically classifies EC into the following subgroups (table 1): endometrioid carcinoma, serous carcinoma, clear cell carcinoma, mixed carcinoma, undifferentiated/dedifferentiated carcinoma, carcinosarcoma, and rare EC subtypes [2].

The WHO also recommends the integration of molecular characteristics to further categorize these tumors and to help predict prognosis. This is discussed in detail below. (See 'Integration of molecular subtypes' below.)

FIGO categorization — The International Federation of Gynecology and Obstetrics (FIGO) 2023 staging system also divides EC into two categories: aggressive and less aggressive tumors [3]. Details of the 2023 FIGO staging system are presented in the tables (table 2 and table 3) and discussed in detail separately. (See "Overview of resectable endometrial carcinoma", section on 'Histopathology' and "Endometrial carcinoma: Staging and surgical treatment", section on 'Staging system'.)

The FIGO 2023 guidelines also recommend the integration of molecular subtypes, when feasible. (See 'Integration of molecular subtypes' below and "Overview of resectable endometrial carcinoma", section on 'Molecular factors'.)

INTEGRATION OF MOLECULAR SUBTYPES

Clinical importance — Molecular classification has been integrated into risk stratification systems and expert clinical guidelines, including the World Health Organization (WHO), European Society of Gynaecological Oncology (ESGO), European Society for Radiotherapy and Oncology (ESTRO), European Society of Pathology (ESP), European Society of Medical Oncology (ESMO), National Comprehensive Cancer Network (NCCN), and International Federation of Gynecology and Obstetrics (FIGO) [2-6].

Incorporating the molecular subtype offers more consistent categorization of ECs and provides both predictive and prognostic information [3,7], and a framework for research endeavors [8-18] and clinical trials [19-25]. The prognostic value of molecular classification has consistently been demonstrated, with predictive value emerging with respect to response to radiotherapy [26,27], chemotherapy [20,24,25,28], and targeted treatment [20,22,29-33]. It is imperative that clinicians are familiar with the key molecular features that enable EC classification, are able to interpret molecular results provided in pathology reports (see 'Pathology reporting' below), and are aware of the implications these features have on clinical care [34].

Molecular subtype assignment can be determined from diagnostic endometrial biopsies or curettings, showing high concordance with classification performed on the hysterectomy specimen [15,35-37]. Approximately 3 percent of ECs have more than one molecular feature [17].

Subtypes — In 2013, The Cancer Genome Atlas (TCGA) used genomic, transcriptomic, and proteomic analyses to characterize over 370 ECs, identifying four molecular subtypes based on tumor cell genomic architecture with distinct prognostic outcomes [7] and clinicopathologic features (table 4). Subsequently, a clinically applicable molecular classification system that can be performed on standard formalin-fixed, paraffin-embedded material and serve as a surrogate for diagnosis of the four TCGA molecular subtypes was developed [7,38-41]. This algorithm arose independently from two research teams, and subsequently through international collaboration has clarified several critical features: consensus on the nomenclature and color designations for the four molecular subtypes [2], and characterization of "multiple classifier" ECs that harbor more than one key classification feature (eg, a tumor with both a POLE mutation and p53 abnormality and/or mismatch repair deficiency). The order of segregation shown in the figure (figure 2) reflects the now agreed upon categorization of these rare tumors.

The four molecular subtypes of EC are as follows, with WHO nomenclature given [42]:

●DNA polymerase epsilon (POLE)-mutated subtype (POLEmut; TCGA "POLE [ultramutated]," previously "POLE," or "POLE EDM") – These are copy number (CN) stable ECs with recurrent mutations in the exonuclease domain of POLE, a gene involved in DNA replication and repair [43-46]. In-depth characterization of ECs with POLE mutations, both within and outside the exonuclease domain, has generated a structured scoring system that limits the assignment of POLEmut to a list of 11 established pathogenic mutations [18]. These tumors have one of the highest somatic mutation frequencies of any solid tumors, frequently exceeding 100 mutations per megabase (Mb). Often, but not exclusively, of endometrioid histologic type, POLEmut ECs have prominent tumor-infiltrating lymphocytes (TILs).

Patients with POLEmut ECs tend to be younger and thinner, and despite often having seemingly aggressive pathologic features (eg, high-grade, lymphovascular space invasion), they have highly favorable outcomes (>96 percent five-year survival) confirmed across multiple studies [47-51].

In an individual patient data meta-analysis of all reported POLEmut ECs, adjuvant therapy was not associated with improved outcomes for women with pathogenic POLE mutations, supporting de-escalation of therapy in clinical trials [51]. Two prospective studies are assessing the possibility of de-escalation of therapy in EC: (1) PORTEC-4a is a multicenter randomized phase III trial in patients with high-intermediate risk EC that has completed accrual but has not yet reported results [52-54], and (2) Tailored Adjuvant Therapy in POLE-mutated and p53-wildtype/no specific molecular profile (NSMP) Early Stage Endometrial Cancer (RAINBO Blue and TAPER) is a prospective cohort study in early-stage EC [51,54].

Immunotherapy may be an option in the rare scenario of recurrent POLEmut EC given the high observed TIL [55-57].

●Mismatch repair-deficient subtype (MMRd; TCGA "MSI [hypermutated]," previously "MMR-D") – These EC have low levels of somatic CN alterations but a very high mutational burden and high TIL secondary to dysfunctional mismatch repair (MMR) proteins (mutL homolog 1 [MLH1], postmeiotic segregation 2 [PMS2], mutS homolog 2 [MSH2], or mutS homolog 6 [MSH6]) [58]. Epigenetic silencing of MLH1 is responsible for the majority of this subtype, but it also includes both somatic and germline mutations (Lynch syndrome (see "Lynch syndrome (hereditary nonpolyposis colorectal cancer): Screening and prevention of endometrial and ovarian cancer")) in any of the MMR genes. This molecular subtype is associated with a high tumor mutational frequency (>10 mutations/Mb).

The receptor tyrosine kinase (RTK)/RAS/beta-catenin pathways and phosphatase and tensin homolog (PTEN)/phosphoinositide 3-kinase (PI3K) signaling pathways are commonly involved in "microsatellite instability" (MSI) ECs (69.5 and >90 percent of cases, respectively) [7,59]. Studies on MSI status and clinical outcomes in EC have had discordant results. Retrospective data suggest increased disease-specific survival in stage IB/II high-grade MMRd endometrioid EC treated with radiation [26], but similar recurrence-free survival was found in a combined cohort of early stage intermediate and high-intermediate risk MMRd EC [27]. The US Food and Drug Administration (FDA) has approved the use of immune checkpoint inhibitors for metastatic or recurrent MSI ECs with marked response rates observed even in the setting of heavily pretreated disease [60,61]. Two landmark clinical trials demonstrated improvement in outcomes in patients with primary advanced or first recurrence ECs treated with immune checkpoint inhibitors in combination with standard chemotherapy [62,63].

●No specific molecular profile (NSMP; TCGA copy-number low [endometrioid], previously also termed p53 wild-type ["p53wt"] and "NSMP/p53wt") – This third molecular subtype is genomically stable, MMR proficient, with moderate mutational load ECs (frequently involving PI3K/Akt and Wnt/catenin beta 1 [CTNNB1] signaling pathways), and has intermediate to favorable outcomes. NSMP ECs have normal p53 immunohistochemical expression and are POLE wildtype. This group encompasses mostly endometrioid neoplasms with estrogen and progesterone receptor (ER, PR) positivity and high response rates to hormonal therapy. Endometrioid ECs of NSMP subtype that are ER negative have been shown to have an unfavorable prognosis [64,65].

●p53 abnormal (p53abn; TCGA copy number [CN] high [serous-like]) – The fourth molecular subtype has high somatic CN alterations and mutational profiles similar to high-grade serous ovarian and basal-like breast carcinomas. TP53 mutations are characteristic for this group. The p53abn cases are associated with a poor prognosis and responsible for 50 to 70 percent of EC mortality. The approximate proportions of p53abn ECs for each histologic type are as follows: serous carcinoma (93 percent), carcinosarcoma (85 percent), clear cell carcinoma (38 percent), grade 3 endometrioid carcinoma (22 percent), and grade 1 or 2 endometrioid carcinoma (5 percent) [66].

Data from patients enrolled in the PORTEC-3 trial and in international cohorts outside of clinical trials suggest that patients with p53abn ECs treated with chemotherapy in addition to radiation have superior outcomes as compared with those who received radiation [25,67].

Human epidermal growth factor receptor 2 (HER2) amplification has been reported in approximately 20 to 25 percent of CN high/p53abn ECs and successfully targeted in serous carcinomas [30,68,69]. Over 40 percent of CN high/p53abn ECs have homologous recombination deficiency (HRD) based on RAD51 foci formation, with a lower percentage showing HRD based on mutational signatures [29,70] presenting another potential target for this molecular subtype. Breast cancer susceptibility genes (BRCA1/2) mutation carriers appear to be at increased risk for p53abn EC, with the highest risk within BRCA1 mutation carriers [71]. Finally, antiangiogenic agents, alone or in combination appear to add value in the treatment of advanced or recurrent p53abn EC [31,72,73].

RISK STRATIFICATION — Multiple risk stratification systems for EC have been developed [19,74-78]. Distinguishing between patients who may be cured by surgery alone and those at significant risk of both local and distant recurrence, and therefore in need of adjuvant therapy, remains a tremendous challenge for clinicians caring for patients with EC.

Risk stratification and approach to adjuvant therapy are discussed in detail separately. (See "Overview of resectable endometrial carcinoma", section on 'Role of adjuvant therapy'.)

HISTOGENESIS AND PRECURSOR LESIONS — ECs that arise from atypical endometrial hyperplasia/endometrioid intraepithelial neoplasia (AEH/EIN) can be associated with any of the four molecular subtypes described above; POLE mutations and loss of mismatch repair (MMR) expression are typically seen in AEH/EIN associated with POLEmut and MMR-deficient (MMRd) EC, respectively, indicating that these alterations are early events in oncogenesis. Serous EC and carcinosarcoma mostly arise from serous endometrial intraepithelial carcinoma (SEIC), but some arise from AEH/EIN (and are associated with mutations in phosphatase and tensin homolog [PTEN]).

●Atypical endometrial hyperplasia/endometrioid intraepithelial neoplasia is characterized by a clonal proliferation of complex endometrial glands such that glands predominate over stroma. The glandular epithelial cells have mild or moderate cytologic atypia, with nuclear stratification and enlargement, sometimes with prominent nucleoli.

AEH/EIN is associated with unopposed estrogenic stimulation and acquisition of mutations in PTEN, Kirsten rat sarcoma viral oncogene homolog (KRAS), phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PI3KCA), PI3K/Akt and Wnt/catenin beta 1 (CTNNB1), and/or AT-rich interaction domain 1A (ARID1A). It gives rise to MMRd and POLEmut ECs, some of the p53-mutated ECs, as well as no specific molecular profile (NSMP) ECs. In an individual patient, the mutational/molecular profiles of AEH/EIN and the concurrent EC are highly concordant [35,79,80].

●Serous endometrial intraepithelial carcinoma – In SEIC, the preexisting endometrial glands are lined by markedly atypical glandular epithelial cells. These epithelial cells show nucleomegaly; typically are less polarized; have rounded, more atypical and pleomorphic nuclei and high nucleus-to-cytoplasm (n:c) ratios; and are associated with a high proliferative index.

Tumor protein 53 (TP53) mutations are the defining feature of SEIC, and mutations in protein phosphatase 2, structural/regulatory subunit alpha (PPP2R1A) are common. SEIC is the precursor of many p53abn ECs, but the percentage is not known.

HISTOLOGIC TYPES — The World Health Organization (WHO) histologically classifies EC into the following subgroups (table 1). (See 'WHO categorization' above.)

Endometrioid carcinoma — Endometrioid EC is the most common EC histology, accounting for 75 to 80 percent of cases.

●Histopathology – Endometrioid EC is composed of tall columnar cells lining back-to-back glands without intervening stroma (suggesting invasion). The glands have a smooth, luminal contour. Cribriform (gland within a gland) patterns are also common. Occasionally, endometrioid ECs have a prominent papillary or villoglandular growth pattern.

Several variant types of endometrioid EC are recognized, including carcinomas with squamous differentiation, mucinous differentiation (characterized by prominent intracellular mucin within the neoplastic cells), and prominent cytoplasmic vacuoles (secretory carcinoma).

Myometrial invasion, if present, may have either an infiltrative pattern with reactive stromal fibrosis and inflammatory cells ("desmoplastic reaction") or a more subtle pushing pattern (large nests of neoplastic cells extend into the myometrium without eliciting a stromal reaction). An unusual pattern seen with grade 1 endometrioid ECs, the microcystic elongated and fragmented pattern ("MELF"), is associated with lymphovascular space invasion and nodal metastases [81,82].

●Grade – Endometrioid ECs are graded using the International Federation of Gynecology and Obstetrics (FIGO) classification system (see 'FIGO categorization' above), which assesses the architectural pattern and nuclear grade [3]:

•Grade 1 – Less than 5 percent solid growth patterns (picture 1)

•Grade 2 – 6 to 50 percent solid growth patterns

•Grade 3 – Greater than 50 percent solid growth (picture 2)

Squamous differentiation is common and not included in assessment of solid growth when determining grade. Marked nuclear atypia (nuclear grade 3) raises the grade by one (from 1 to 2 or from 2 to 3) [3].

●Molecular subtypes – Endometrioid carcinomas can be any of the four molecular subtypes (see 'Integration of molecular subtypes' above), which indicates the genetic heterogeneity within this histologic type. Many grade 3 endometrioid ECs have a genomic profile similar to serous ECs, with high copy number (CN) alterations and tumor protein 53 (TP53) mutations [83-85], and represent a particularly diverse subset of EC [16].

●Prognosis – Most endometrioid ECs are low grade (grade 1 or 2), diagnosed at an early stage, and have a good prognosis [86]. Approximately 5 percent of low-grade endometrioid carcinomas are p53abn with worse outcomes [83,87]. Grade 3 p53abn endometrioid ECs behave more aggressively and have a poorer prognosis than other molecular subtypes of grade 3 endometrioid EC [16]. (See "Treatment of low-risk endometrial cancer", section on 'Prognosis'.)

Serous endometrial carcinoma — Serous endometrial carcinoma (SEC) is the second most common type of EC but only accounts for approximately 10 percent of cases.

●Histopathology – In SEC, the neoplastic cells form papillary structures and glands with serrated outlines. The cells have marked nuclear atypia with prominent nucleoli and numerous mitotic figures (picture 3). Occasionally, SEC is composed primarily of glands, but the marked nuclear atypia is a key to proper tumor classification. Psammoma bodies may be present.

●Molecular subtypes – A very large majority of SECs are of the p53abn molecular subtype (see 'Integration of molecular subtypes' above). Human epidermal growth factor receptor 2 (HER2) is overexpressed/amplified/mutated in a minority of SECs and can be targeted therapeutically [30].

●Prognosis – Clinically occult extrauterine disease is often present at diagnosis [88]. SEC often diffusely infiltrates the myometrium and may have extensive lymphovascular space invasion and peritoneal spread, similar to ovarian carcinoma. However, SEC confined to the endometrium (or a polyp) with minimal myometrial invasion and no distant disease after surgical staging has a good prognosis.

SEC is discussed in more detail separately. (See "Endometrial carcinoma: Serous and clear cell histologies".)

Clear cell carcinoma — Clear cell EC is an uncommon subtype, comprising <5 percent of EC, and patients are usually older and postmenopausal [89,90].

●Histopathology – Clear cell ECs share morphologic features with ovarian clear cell carcinoma and have several different architectural patterns: papillary, glandular, tubulocystic, and diffuse. The neoplasm is often composed of cells with abundant clear cytoplasm (picture 4); however, eosinophilic cytoplasm, hobnail, and flat cells can also be seen. Cytologic atypia is variable, and occasional enlarged irregular nucleoli are seen.

●Molecular subtypes – Clear cell ECs can be of any of the four molecular subtypes (see 'Integration of molecular subtypes' above); POLEmut clear cell carcinomas have the most favorable prognosis while p53abn clear cell carcinomas are associated with aggressive behavior [11,14,91]. The mismatch repair-deficient (MMRd) cases often show mixed morphology, with clear cell and endometrioid components [92].

Clear cell ECs are typically negative for estrogen receptor (ER) protein and positive for Napsin A, which can aid in distinguishing this form of high-grade carcinoma from its mimics: SEC and the secretory variant of endometrioid EC.

The 2020 European Society of Gynaecological Oncology (ESGO), European Society for Radiotherapy and Oncology (ESTRO), and European Society of Pathology (ESP) guidelines do not offer specific recommendations for cases of clear cell EC that have undergone molecular classification, citing insufficient evidence to direct risk group assignment and treatment in these rare cancers [4].

Clear cell endometrial carcinoma is discussed in more detail separately. (See "Endometrial carcinoma: Serous and clear cell histologies".)

Mixed carcinoma — Mixed carcinomas have at least two distinct histologic components, typically endometrioid and a high-grade nonendometrioid pattern (usually serous, sometimes clear cell). These neoplasms are almost all clonal rather than being a collision between two synchronous but independent primary neoplasms and are usually the same molecular subtypes throughout (ie, they are an example of morphologic variability within a molecular subtype).

Undifferentiated/dedifferentiated carcinoma — Undifferentiated/dedifferentiated ECs are the least well understood of the major histologic types of EC [93-95].

These neoplasms have no glandular or squamous differentiation. Most express epithelial antigens (eg, cytokeratin), but this is typically focal. Dedifferentiated carcinomas are composed of International Federation of Gynecology and Obstetrics (FIGO) grade 1 or 2 endometrioid EC adjacent to areas of undifferentiated carcinoma. Undifferentiated/dedifferentiated carcinomas are frequently MMRd, often have mutations in genes encoding proteins of the switch/sucrose nonfermentable (SWI/SNF) complex, and, in most cases, do not have mutations in TP53.

Carcinosarcoma — Carcinosarcoma (previously known as malignant mixed Müllerian tumor) is an uncommon, aggressive, biphasic carcinoma (not sarcoma) that accounts for <5 percent of ECs.

●Histopathology – Carcinosarcoma is composed of a high-grade sarcoma component juxtaposed with a high-grade carcinoma. The sarcomatous component is composed of cell types intrinsic to the uterus (homologous tumors), such as endometrial stromal sarcoma or leiomyosarcoma, or cell types extrinsic to the uterus (heterologous tumors), such as chondrosarcoma or rhabdomyosarcoma. The carcinomatous component is usually high grade, frequently difficult to assign to a specific histologic type, and can show features of high-grade endometrioid carcinoma, serous carcinoma, clear cell carcinoma, or undifferentiated carcinoma (picture 5).

Metastatic neoplasms are purely carcinomatous or mixed carcinoma and sarcoma in over 90 percent of cases with only a few comprising pure sarcoma [96].

●Molecular subtypes – A majority of carcinosarcomas are of the p53abn molecular subtype (90 percent of carcinosarcomas characterized by The Cancer Genome Atlas [TCGA] had TP53 mutations), but they can also be of the other three molecular subtypes of EC [97,98]. (See 'Integration of molecular subtypes' above.)

Carcinosarcoma is discussed in more detail separately. (See "Clinical features, diagnosis, staging, and treatment of uterine carcinosarcoma".)

Rare EC subtypes — Mesonephric and mesonephric-like adenocarcinoma of the uterine corpus resemble morphologically their better described counterparts in the uterine cervix. They have a characteristic immunophenotype (GATA-binding protein 3 [GATA3] positive, thyroid transcription factor-1 [TTF1] positive, ER negative) and are associated with aggressive behavior.

Squamous cell carcinomas of the endometrium, in their pure form, are rare and must be distinguished from the much more common endometrioid adenocarcinomas with squamous differentiation, which can be predominantly squamous. They can arise in association with chronic inflammation and squamous metaplasia.

PATHOLOGY REPORTING — The International Collaboration on Cancer Reporting (ICCR) determines the core and noncore factors that should be reported for EC resection specimens and their EC histopathology reporting guide can be found online [99].

The diagnostic report for EC based on an endometrial biopsy or curetting specimen includes neoplasm histology and, for endometrioid carcinomas, grade. The following is a pragmatic approach to use of molecular markers at this time, acknowledging that this is an area in evolution.

●Perform mismatch repair (MMR) immunohistochemistry (or microsatellite instability [MSI] assay) in all EC cases. In most institutions, the most cost-effective approach is to screen with mutL homolog 6 (MSH6) and postmeiotic segregation 2 (PMS2) immunohistochemistry alone, with subsequent mutL homolog 2 (MSH2) or mutL homolog 1 (MLH1) immunohistochemistry when needed. The determination of MMR expression is recommended by the Society of Gynecologic Oncology (SGO) as part of Lynch syndrome screening in EC patients.

●Protein 53 (p53) immunostaining (or molecular testing) should be performed in all high-grade endometrioid and clear cell ECs and in low-grade neoplasms with any unusual features, such as increased nuclear atypia or mitotic activity. The role of p53 immunostaining in "typical" low-grade endometrioid carcinomas is uncertain; most centers do not routinely perform p53 testing of grade 1 or grade 2 EC as relatively few will show abnormal p53 immunostaining. Because patients with p53abn EC, irrespective of histology, will be candidates for adjuvant chemotherapy based on data from PORTEC-3 [25], there is a trend towards increased p53 testing in EC, with reflex p53 immunostaining of all EC performed in some centers.

●POLE mutational testing may be performed for qualification in clinical trials and in clinical practice where adjuvant treatment decisions would be altered. The British Association of Gyneacological Pathologist (BAGP) recommends POLE testing in all stage I/II cases with mismatch repair-deficient (MMRd) and/or p53 abnormalities, nonendometrioid histology, poorly differentiated (grade 3) endometrioid histotype and low-grade endometrioid histology together with high-risk factors such as lymphovascular space invasion, deep myometrial invasion, and estrogen receptor (ER) negativity [100]. Subclonal abnormal p53abn immunostaining could also be an indication as approximately one-half of ECs with subclonal complete loss or overexpression of p53 are POLEmut and thus associated with a favorable (POLEmut) rather than unfavourable (p53abn) EC molecular subtype.

●ER immunohistochemical staining is an important stratification feature within no specific molecular profile (NSMP) ECs and can also be performed on diagnostic biopsy/curettings [64,65] to inform prognosis, guide treatment decisions, and determine which individuals need POLE testing [100].

●Human epidermal growth factor receptor 2 (HER2) overexpression or amplification can be assessed by immunohistochemistry or in situ hybridization, respectively, and can provide both prognostic and predictive value within serous/p53abn ECs [30,69,101].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Uterine cancer".)

SUMMARY AND RECOMMENDATIONS

●Classification systems – The classification of endometrial cancers (ECs) has evolved over time.

•Traditional – ECs have traditionally been classified as type I or type II. Type I ECs are estrogen driven, mostly endometrioid histology, lower grade, typically have less myometrial invasion than type II EC, and are associated with a favorable prognosis. By contrast, type II ECs comprise a diverse mix of high-grade, clinically aggressive histologies (eg, serous, clear cell), with a poor response rate to progestogens and a poor outcome. (See 'Traditional classification system' above.)

•WHO – The World Health Organization (WHO) histologically classifies EC into the following subgroups (table 1): endometrioid, serous, clear cell, mixed, and undifferentiated/dedifferentiated carcinoma, carcinosarcoma, and rare EC subtypes. (See 'WHO categorization' above.)

•FIGO – The International Federation of Gynecology and Obstetrics (FIGO) 2023 staging system divides histological types into aggressive and less aggressive tumors (table 2 and table 3); this is described in detail separately. (See 'FIGO categorization' above and "Overview of resectable endometrial carcinoma", section on 'Histopathology'.)

●Integration of molecular subtypes – Molecular features and subtyping are integrated into pathology reporting, where available, offering more consistent categorization of ECs and providing both predictive and prognostic information. The four molecular subtypes of EC are POLE mutant (POLEmut), mismatch repair-deficient (MMRd), no specific molecular profile (NSMP), and p53 abnormal (p53abn) (table 4). (See 'Integration of molecular subtypes' above and 'Pathology reporting' above.)

●Precursor lesions – ECs that arise from endometrial hyperplasia with atypia/endometrioid intraepithelial neoplasm (AEH/EIN) can be associated with any of the four molecular subtypes of EC; POLE mutations and mismatch repair (MMR) loss of expression are typically seen in AEH/EIN associated with POLEmut and MMRd EC, respectively, indicating that they are early events in oncogenesis. Serous EC and carcinosarcoma mostly arise from serous endometrial intraepithelial carcinoma, but some arise from AEH/EIN (and are associated with mutations in phosphatase and tensin homolog [PTEN]). (See 'Histogenesis and precursor lesions' above.)

●Pathology reporting – The International Collaboration on Cancer Reporting (ICCR) determines the core and noncore factors that should be reported for EC resection specimens, and their EC histopathology reporting guide can be found online. (See 'Pathology reporting' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Lesley Lomo, MD; Jonathan Hecht, MD, PhD; and Margaret Steinhoff, MD, who contributed to earlier versions of this topic review.