Background
The Cancer Genome Atlas comprehensive molecular analysis, N=373 endometrial carcinomas (endometrioid and serous). Integrated whole-exome sequencing, MSI analysis, copy number, mRNA/miRNA, and protein analysis to define clinically significant molecular subtypes.
Interventions and follow up
Treatment: TCGA integrated genomic, transcriptomic, proteomic, and clinical analysis of 373 endometrial carcinomas
Primary endpoint: Identification of molecular subgroups (POLE ultramutated, MSI hypermutated, copy-number low, copy-number high/serous-like)
mFollow up: Not applicable
Primary endpoint: Identification of molecular subgroups (POLE ultramutated, MSI hypermutated, copy-number low, copy-number high/serous-like)
mFollow up: Not applicable
Results
POLE ultramutated (7%): Excellent prognosis regardless of grade; predominantly endometrioid; POLE exonuclease-domain mutations; very high tumor mutational burden
MSI hypermutated (28%): MLH1 hypermethylation most common; intermediate prognosis; predominantly endometrioid grade 1–2; high immunotherapy sensitivity
Copy-number low (39%): Predominantly endometrioid grade 1–2; CTNNB1 mutations common; intermediate-good prognosis
Copy-number high/serous-like (26%): TP53 mutation; predominantly serous and high-grade endometrioid; worst prognosis
MSI hypermutated (28%): MLH1 hypermethylation most common; intermediate prognosis; predominantly endometrioid grade 1–2; high immunotherapy sensitivity
Copy-number low (39%): Predominantly endometrioid grade 1–2; CTNNB1 mutations common; intermediate-good prognosis
Copy-number high/serous-like (26%): TP53 mutation; predominantly serous and high-grade endometrioid; worst prognosis
Adverse events
Toxicity: Not applicable
Type: Molecular genomic analysis, not a treatment study
Type: Molecular genomic analysis, not a treatment study
Conclusions
TCGA identified 4 molecular subtypes of endometrial cancer with distinct genomic profiles, prognostic implications, and therapeutic vulnerabilities, providing the scientific foundation for molecular-based adjuvant therapy decisions (e.g., PORTEC-4, RAINBO).
Key Limitations
Descriptive molecular characterization, not a prospective therapeutic trial; full TCGA classifier requires sequencing not always available in routine practice (surrogate ProMisE/molecular classifiers used clinically); prognostic associations not derived from randomized therapy data.
Clinical Context
Defined the molecular taxonomy now embedded in ESMO/ESGO endometrial classification and guiding adjuvant therapy (including radiotherapy de-escalation/escalation) in molecular-stratified trials; surrogate classifiers (e.g., ProMisE) approximate TCGA groups in practice.