How is endometrial cancer classified?

Endometrial Cancer Classification

Endometrial cancer is now classified using an integrated molecular system that divides tumors into four distinct prognostic groups: POLE-ultramutated (POLEmut), mismatch repair deficient (dMMR), no specific molecular profile (NSMP), and p53-abnormal (p53-abn), which has replaced the outdated type I/type II histopathological classification. 1

The Four Molecular Subgroups

The Cancer Genome Atlas (TCGA) molecular classification stratifies endometrial cancer based on mutational burden and copy number alterations, providing superior prognostic information compared to histology alone 1:

1. POLE-Ultramutated (POLEmut)

• Prevalence: 5-15% of cases 1
• Molecular features: >100 mutations/megabase, very low somatic copy number alterations, microsatellite stable 1
• Pathogenic variants: P286R, V411L, S297F, A456P, S459F 1
• Diagnostic test: Next-generation sequencing, Sanger sequencing, or hotspot mutation analysis 1
• Histology: Frequently endometrioid with high-grade features, ambiguous morphology, prominent tumor-infiltrating lymphocytes and tertiary lymphoid structures 1
• Clinical features: Lower BMI, early stage (IA-IB), early onset 1
• Prognosis: Excellent—the most favorable outcome despite often presenting as high-grade tumors 1, 2

2. Mismatch Repair Deficient (dMMR/MSI)

• Prevalence: 25-30% of cases 1
• Molecular features: 10-100 mutations/megabase, low somatic copy number alterations, microsatellite instability 1
• Diagnostic test: Immunohistochemistry for MLH1, MSH2, MSH6, PMS2 (loss of one or more proteins indicates dMMR) 1
• Histology: Often endometrioid with high-grade features, substantial lymphovascular space invasion, prominent tumor-infiltrating lymphocytes 1
• Clinical features: Higher BMI, associated with Lynch syndrome 1
• Prognosis: Intermediate 1

3. No Specific Molecular Profile (NSMP/p53-wild type)

• Prevalence: 30-40% of cases 1
• Molecular features: <10 mutations/megabase, low somatic copy number alterations, microsatellite stable 1
• Diagnostic test: Exclusion diagnosis after ruling out POLE mutations, dMMR, and p53 abnormalities 1
• Histology: Mostly low-grade endometrioid with squamous differentiation, notable absence of tumor-infiltrating lymphocytes, diffuse ER/PgR expression 1
• Clinical features: Higher BMI 1
• Prognosis: Intermediate 1

4. p53-Abnormal (p53-abn/Copy Number High)

• Prevalence: 5-15% of cases 1
• Molecular features: <10 mutations/megabase, high somatic copy number alterations, microsatellite stable 1
• Diagnostic test: p53 immunohistochemistry (abnormal pattern indicates TP53 mutation) 1
• Histology: All histological subtypes including serous, carcinosarcoma; mostly high-grade with high cytonuclear atypia, low tumor-infiltrating lymphocytes 1, 2
• Clinical features: Lower BMI, advanced stage, late onset 1
• Prognosis: Poor—the worst outcome, almost always requires chemotherapy 1, 2

Diagnostic Algorithm

The ESMO guideline provides a hierarchical testing approach 1:

1. First: Test for POLE hotspot mutations (P286R, V411L, S297F, A456P, S459F) in appropriate cases 1
2. Second: Perform mismatch repair immunohistochemistry (MLH1, PMS2, MSH2, MSH6) 1
3. Third: In mismatch repair-proficient, POLE wild-type tumors, perform p53 immunohistochemistry 1
4. Classification: Tumors are assigned hierarchically—POLEmut supersedes all others, then dMMR, then p53-abn, with NSMP as the default when all tests are negative/normal 1

Traditional Histopathological Classification (Still Reported)

While molecular classification is now preferred, histological subtypes remain relevant 1, 2:

• Endometrioid carcinoma: Comprises approximately 80% of cases, characterized by glandular architecture, graded by solid growth pattern (Grade 1: ≤5%, Grade 2: 5-50%, Grade 3: >50% solid growth) 2, 3
• Serous carcinoma: Complex papillary architecture, marked nuclear atypia, predominantly p53-abnormal molecular subtype, 70% present with extrauterine spread 2
• Clear-cell carcinoma: Rare, high-risk histology 1
• Carcinosarcoma: Now recognized as metaplastic carcinoma with molecular overlap with serous and endometrioid types, should be molecularly classified 1, 3
• Undifferentiated/dedifferentiated carcinoma: High-risk histology 1

Transition Period Nomenclature

During this transition from histology-based to molecular classification, specify which system is used 1:

• Tumors not molecularly classified should be designated "EC not-otherwise-specified (EC-NOS)" 1
• Example: "endometrioid-type EC (EEC-NOS)" for tumors lacking molecular testing 1
• This notation clarifies for clinicians and patients which classification framework applies 1

Clinical Implications

The molecular classification directly impacts treatment decisions 1:

• POLEmut tumors: May avoid adjuvant therapy even when high-grade or deeply invasive, currently being tested in PORTEC-4a trial 1
• p53-abn tumors: Derive greatest benefit from adjuvant chemotherapy regardless of stage, includes most serous carcinomas and significant portions of other histologies 1
• dMMR tumors: Potential candidates for immunotherapy 1
• NSMP tumors: Standard risk-adapted therapy based on stage and grade 1

Critical Pitfalls to Avoid

• Do not rely solely on histology to determine prognosis—a high-grade endometrioid tumor may be POLEmut with excellent prognosis, or p53-abn with poor prognosis 1
• Do not skip molecular testing in apparent early-stage disease, as p53-abn tumors require aggressive treatment regardless of stage 1
• Do not assume all serous carcinomas are p53-abn—approximately 30% are not, and molecular classification provides more accurate prognostication 1
• Perform hierarchical testing in the correct order (POLE → MMR → p53) to avoid misclassification 1