Uterus

• Malignant epithelial neoplasm originating from the endometrium

• From a pathophysiologic perspective, endometrial carcinomas have been traditionally divided into 2 types:
  • Type 1: includes endometrioid and mucinous carcinoma
    • These lesions are associated with long term elevated estrogen levels, which lead to persistent proliferative stimulation of the endometrium
    • Risk factors leading to hyperestrogenism include obesity, exogenous hormonal therapy (e.g. tamoxifen use for breast cancer), ovarian cortical hyperplasia / hyperthecosis, polycystic ovarian syndrome and hormone producing tumors (e.g. granulosa cell tumor of the ovary), among others
    • PTEN, ARID1A, PIK3CA, KRAS gene alterations are common
    • Atypical endometrial hyperplasia / endometrioid intraepithelial neoplasia is regarded as the precursor lesion (J Clin Oncol 2010;28:788, Cancer 2005;103:2304)
  • Type 2: includes serous, clear cell, undifferentiated carcinoma and carcinosarcoma
    • These tumors have a lesser association with unopposed estrogen exposure
    • Serous carcinoma is characterized by early alterations in TP53
    • Precursor lesion for clear cell carcinoma has not been identified
• Other associations include diabetes, dysfunctional uterine bleeding, hypertension, infertility, Muir-Torre syndrome, Turner syndrome (usually well differentiated adenocarcinoma; 2/3 have squamous differentiation) and tamoxifen use for breast cancer (increased risk for endometrioid, serous carcinoma and carcinosarcoma) (Am J Surg Pathol 2001;25:936, OMIM: Muir-Torre Syndrome [Accessed 1 May 2020])
• From a biologic and clinical perspective, the classification of endometrial carcinoma is evolving towards a molecular based grouping (see Molecular / cytogenetics description below)

Contributed by Carlos Parra-Herran, M.D.

• Distinct mass or growth is usually seen upon examination of the cavity
  • Lesion is typically exophytic and soft
  • It can have homogeneous appearance or contain a heterogeneous cut surface with variable hemorrhage and necrosis (more common in high grade tumors)
• Some cases present as diffuse endometrial thickening; the endometrial tissue is lush, soft and friable
• Carcinomas with extensive squamous differentiation can have a flaky appearance, whereas those with mucinous differentiation are soft and gelatinous (colloid appearance)
• Tumor should be sectioned perpendicular to the anterior and posterior uterine walls in order to identify areas of growth into the wall
  • These include myometrial invasion and involvement of adenomyosis, which may be difficult to distinguish grossly
  • Area of deepest growth into the wall should always be sampled as a full thickness section (either a single section or a composite section if the uterine wall is thick)
  • Inclusion of the adjacent uninvolved endometrium in the section is always preferred (if present)
• If the tumor grossly extends to lower uterine segment, a sagittal section, going from proximal to distal ends to include tumor, lower uterine segment and cervix is recommended
• References: Lester: Manual of Surgical Pathology, 3rd Edition, 2010, Nucci, Parra-Herran: Gynecologic Pathology, 2nd Edition, 2020

• See endometrial carcinoma types

• Pathologic definition of carcinoma of the endometrium:
  • Diagnosis of carcinoma is based on features indicative of invasion into the surrounding mesenchyme (endometrial stroma or myometrium)
    • Stromal invasion is typically seen in the form of glandular confluence and complex architecture: loss of individual glandular contours with gland fusion, lack of intervening stroma and back to back architecture
    • Invasion usually presents in cribriform, microacinar and solid architectural patterns; other helpful features are stromal desmoplasia (stroma has myofibroblasts, edema, inflammatory cells and myxoid change), stromal necrosis (stroma replaced by necrotic and inflammatory debris) or combinations of these findings between adjacent glands
    • So called microcystic, elongated and fragmented (MELF) pattern of growth represents, by definition, invasion
      • This pattern can be deceptive and easily missed at low power magnification
• Depth of myometrial invasion, cervical stromal involvement, serosal involvement and adnexal involvement are relevant to pT category (see staging)
• Histologic grading:
  • Endometrioid and mucinous carcinomas are graded with a 3 tier system developed by the International Federation of Gynecology and Obstetrics (FIGO):
    • FIGO 1: predominant glandular growth and < 5% nonsquamous solid component; glandular architecture is identified by the presence of patent lumina within the gland, relatively preserved polarity of the epithelium and absent to mild epithelial stratification
    • FIGO 2: 6 - 50% nonsquamous solid component
    • FIGO 3: > 50% nonsquamous solid component
    • Architectural grading described above is upgraded by 1 if there is severe nuclear atypia (pleomorphism, nuclear enlargement and nucleoli evident at low power magnification)
      • Endometrioid carcinoma FIGO grade 2 purely based on cytologic atypia (that is, with severe atypia but architecturally well differentiated) is extremely rare and must be treated as a diagnosis of exclusion; it is imperative to first exclude serous and clear cell carcinoma
    • In general, a 2 tier system can be also applied, with FIGO1 and FIGO2 being considered low grade and FIGO 3 being considered high grade
  • Other carcinoma types (serous, clear cell, carcinosarcoma, undifferentiated, mixed) are by definition high grade
• Lymph node involvement:
  • 2018 AJCC staging manual introduces different categories for lymph node metastases in gynecologic malignancies, mostly based on size
• Lymphovascular space invasion:
  • Lymphovascular space invasion (LVI) is an independent predictor of nodal metastases and local recurrence (Gynecol Oncol 2012;124:31, Arch Gynecol Obstet 2013;288:1391)
  • LVI is defined as tumor cells in a space lined by endothelial cells outside the immediate invasive border
  • Extent of LVI, not just the presence, correlates significantly with regional and distal lymph node involvement, locoregional recurrence and survival (Eur J Cancer 2015;51:1742, Histopathology 2019;75:128)
    • Extent should be reported as follows:
      • Absent: no LVI as defined
      • Focal: a single focus of LVI
      • Substantial: diffuse or multifocal LVSI
  • Under this system, the term focus is understood as a cluster containing up to 5 individual involved vascular spaces
    • It can be inferred that substantial LVI represents either > 1 focus as defined or any focus with > 5 individual involved vascular spaces
  • Artificial tumor intrusion into vascular spaces (vascular pseudoinvasion) can be seen in laparoscopic, robotic assisted hysterectomies (Am J Surg Pathol 2008;32:560, Am J Surg Pathol 2009;33:298)
    • This phenomenon occurs not only in cancer related surgery: displacement of normal endometrial glands and stroma has been reported in 13% of laparoscopic hysterectomies performed for benign conditions (Am J Surg Pathol 2008;32:560)
  • Of note, other studies have shown no association between laparoscopic hysterectomy or the use of a uterine manipulator and the prevalence of LVI (JSLS 2014;18:e2014.00021, Am J Obstet Gynecol 2013;208:71.e1)
  • Real vascular invasion is seen in lymphatics and venous vessels, not in arterial vessels; intravascular foci are round and conform to the shape of the vessel; sometimes they are partially adherent to the vessel wall; cells have more eosinophilic cytoplasm and rounder shape compared with the native tumor; the presence of a perivascular lymphocytic infiltrate also supports real LVI (Mod Pathol 2010;23:1073)
  • Artificial intrusion should be considered if the intravascular tumor retains a gland shape or stromal elements within it, has a large ("chunky") size and involves arterial vessels or other elements are identified within vascular spaces (benign endometrium, surface necrotic material or exudates) (Mod Pathol 2010;23:1073)

• Abnormalities in cervicovaginal cytology can be the first presenting sign of endometrial carcinoma
• Presence of endometrial cells in pap smears of women over the age of 45 needs to be reported (Cytojournal 2017;14:22)
• Diagnosis of adenocarcinoma in cervicovaginal cytology should prompt consideration for endometrial sampling
• See endometrial carcinoma types for detailed description

• See endometrial carcinoma types

• CK7, CK8 / 18, CK19
• Also vimentin (65%), CEA (areas of squamous metaplasia), CA-125, ER, PR, PTEN (80%), cyclin D1 (40%)
• CD10, as a marker of endometrial stroma, is not useful to distinguish myometrial invasion from adenomyosis involvement (Mod Pathol 2003;16:22, Am J Surg Pathol 2003;27:786)
  • IFITM1, another endometrial stroma marker, has superior specificity in this differential (Am J Clin Pathol 2016;145:486)

• Determination of the tumor histologic type is critical for patient risk stratification and management
• However, there is poor interobserver reproducibility in tumor type and grade among expert pathologists (Mod Pathol 2013;26:1594, Am J Surg Pathol 2013;37:874, Int J Gynecol Cancer 2011;21:654)
• Recent studies have provided a comprehensive characterization of the genomic profiles of endometrial carcinoma:
  • In 2013, The Cancer Genome Atlas (TCGA) Research Network published an integrated genomic characterization of endometrial carcinoma based on genomic data from array and sequencing based technologies (Nature 2013;497:67)
  • It proposed a classification that separates endometrial carcinomas in 4 groups:
    • Copy number - high (frequently involving mutations of TP53); this group includes the vast majority of serous carcinomas and 25% of high grade endometrioid tumors
    • Copy number - low (frequently involving mutations of PTEN, PIK3CA, ARID1A and KRAS); this group is mostly composed of low grade endometrioid carcinomas
    • Microsatellite instability hypermutated (frequently involving alterations of mismatch repair protein genes)
    • Polymerase ε (POLE) ultramutated; this group is mostly composed of endometrioid cancers, which despite having a dramatically increased transversion mutation frequency and newly identified hotspot mutations in the POLE gene (which encodes the central catalytic subunit of DNA polymerase epsilon), appear to have a better prognosis than other groups (J Natl Cancer Inst 2014;107:402, Cancer 2015;121:386)
  • Molecular based classification correlates with clinical outcomes: survival rates are best in POLE mutated tumors, followed by copy number - low, microsatellite instability and copy number - high carcinomas (Nature 20132;497:67)
    • Thus, the molecular fingerprint can better assist in patient risk stratification and management
  • Ancillary testing using formalin fixed, paraffin embedded tumoral tissue can serve as a surrogate to detect its molecular alterations and determine the molecular group (J Pathol 2012;228:20)
  • PROMISE algorithm is based on POLE mutational analysis and immunohistochemistry for p53 and mismatch repair proteins (MLH1, MSH2, MSH6 and PMS2) as a valid surrogate to determine the molecular group (Br J Cancer 2015;113:299, Diagram 2)
    • When combined with clinicopathologic features, the molecular classifier is highly correlated with outcome and survival curves
  • Some carcinomas harbor more than one molecular classifying feature and are referred to as multiple classifier; recent evidence suggests that MMR deficiency and POLE mutations supersede p53 alterations in terms of clinical behavior (J Pathol 2020;250:312)
    • MMR deficient, p53 abnormal tumors should be categorized in the MMR deficient / microsatellite instable group
    • POLE mutant, p53 abnormal tumors should be classified in the ultramutated POLE mutant group
    • POLE mutant, MMR deficient should be classified in the ultramutated POLE mutant group (J Pathol 2020;250:323)
  • CTNNB1 mutations have been found to be an adverse prognostic feature in patients with low grade, low risk endometrial carcinoma (Mod Pathol 2017;30:1032)
  • Nuclear expression of beta catenin is usually associated with CTNNB1 mutations, although the correlation is not perfect (Mod Pathol 2018;31:1553)

Which of the following prognostic variables modifies the FIGO / AJCC stage of endometrial carcinoma?

1. Depth of myometrial invasion
2. Involvement of adenomyosis
3. Lymphovascular space invasion
4. Molecular group
5. Superficial (glandular) cervical involvement

A. Depth of myometrial invasion. All the options have documented prognostic impact in patients with endometrial carcinoma; however, only depth of myometrial invasion modifies the pathologic stage.

Comment Here

Reference: Endometrial carcinoma - general

Which of the following variables modifies the FIGO / AJCC stage of endometrial carcinoma?

1. Extranodal extension by carcinoma involving a lymph node
2. Lower uterine segment involvement
3. MELF pattern of invasion
4. Number of intravascular tumor foci
5. Size of lymph node tumor metastases

E. Size of lymph node tumor metastases. Of all the included features, only the size of the lymph node alters the stage. Tumor metastasis now subclassifies the N (FIGO stage III) category in N0i+ (isolated tumor cells), N1 / N2mi (micrometastases) and N1 / N2a (macrometastases).

Comment Here

Reference: Endometrial carcinoma - general