Table 2

Association between molecular classification and prognosis, adjusted on known prognostic factors

Model 1Univariate analysesMultivariate analyses
Number of patientsNumber of eventsHR (95% CI)*p value*Adjusted
HR (95% CI)†
Adjusted p value†
FIGO stage
 I–II861511
 III–IV35173.26 (1.60 to 6.63)0.0012.35 (1.10 to 5.02)0.029
Histological grade‡
 Low grade901611
 High grade35184.32 (2.16 to 8.61)<0.0011.47 (0.61 to 3.53)0.390
Molecular groups
 Non-TP53-mutated classified§881611
TP53-mutated classified30186.15 (3.00 to 12.6)<0.0015.54 (2.30 to 13.4)<0.001
Model 2Number of patientsNumber of eventsHR (95% CI)*p value*Adjusted
HR (95% CI)¶
Adjusted p value¶
Clinical risk of relapse**
 Low/intermediate/high-intermediate risk721111
 High risk40132.50 (1.09 to 5.71)0.0311.76 (0.73 to 4.23)0.208
Molecular groups
 Non-TP53-mutated classified§831411
TP53-mutated classified22104.71 (2.00 to 11.1)<0.0013.92 (1.59 to 9.64)0.003
  • HR (95% CI): hazard ratio with 95% confidence interval, estimated using the Cox logistic regression model. p value estimated using the Cox logistic regression model.

  • Model 1 included all patients with available data for each variable considered in the model (N=116 patients).

  • Model 2 included only patients without advanced/metastatic disease (N=105 patients), in order to estimate the association between TP53/CNH-like tumors and event-free survival, independently of clinical risk of relapse.

  • *All values displayed are estimated in univariable analysis.

  • †All values displayed are adjusted on the three variables included in the multivariable model 1.

  • ‡Low grade tumors included grade 1 and 2 endometrioid carcinomas. High grade tumors included grade 3 endometrioid carcinomas and type 2 carcinomas.

  • §Non-TP53-mutated classified tumors: belonging to the POLE/ultramutated-like, MSI/hypermutated-like (mismatch repair deficient), and not otherwise specified groups, compared to the TP53-mutated molecular group.

  • ¶All values displayed are adjusted on the two variables included in the multivariable model 2. Model 2 included all patients with available data for each variable considered in the model.

  • **Based on ESMO-ESGO-ESTRO consensus.3 4

  • CNH, copy-number-high; EFS, event-free survival; ESMO-ESGO-ESTRO, European Society for Medical Oncology-European Society of Gynaecological Oncology-European SocieTy for Radiotherapy & Oncology; MSI, microsatellite instability.