Table 1 Prognostic factors for recurrence in high-risk endometrial cancer.

From: Tertiary lymphoid structures critical for prognosis in endometrial cancer patients

Recurrence

Pathologic model

Molecular model

Molecular-immune model

N = 378, 111 events

HR

95% CI

p-value

HR

95% CI

p-value

HR

95% CI

p-value

Age

1.05

1.02–1.07

0.00032

1.03

1.00–1.05

0.045

1.03

1.00–1.05

0.034

Adjuvant treatment

         

RT

reference

 

reference

 

reference

 

CTRT

0.76

0.52–1.10

0.15

0.70

0.48–1.03

0.069

0.72

0.50–1.063

0.093

Histograde

         

EEC grade 1-2

reference

 

reference

 

reference

 

EEC grade 3

1.24

0.76–2.02

0.40

1.23

0.71–2.14

0.45

1.37

0.79–2.37

0.27

non-EEC

1.42

0.89–2.27

0.14

0.94

0.53–1.70

0.85

1.02

0.57–1.82

0.96

Stage (I-II vs. III)

1.93

1.28–2.91

0.0016

1.90

1.27–2.85

0.0020

1.98

1.31–2.97

0.0011

LVSI

1.25

0.82–1.96

0.30

1.29

0.83–2.00

0.25

1.24

0.80–1.94

0.33

Molecular group

         

NSMP

   

reference

 

reference

 

POLEmut

   

no events

 

no events

 

MMRd

   

0.89

0.53–1.48

0.64

0.99

0.59–1.64

0.96

p53abn

   

2.63

1.46–4.73

0.0012

2.42

1.35–4.32

0.0029

TLS

      

0.32

0.14–0.73

0.0073

  1. Three Cox proportional hazards models to showing impact of respectively clinicopathological factors, clinicopathological + molecular factors, and clinicopathological + molecular factors + presence of tertiary lymphoid structures on time to recurrence. Covariates were pre-specified according to Léon-Castillo et al. (J Clin Oncol, 2020)2. The addition of the molecular classifier to the pathologic model was associated with an improvement in model fit evidenced by: (i) reduction in Akaike’s information criterion (AIC) 1244.968 vs. 1202.158, (ii) increase in model concordance (C index 0.655 vs. 0.714, and (iii) likelihood ratio test for comparison of nested models p = 1.43 × 10−10. Likewise, the addition of TLS presence improved model fit: (i) AIC 1202.158 vs. 1194.242, (ii) C index 0.714 vs. 0.729, (iii) likelihood ratio test for nested models p = 0.0016.
  2. HR Hazard ratio; CI Confidence interval; RT Radiotherapy; CTRT Chemoradiation; EEC Endometrioid endometrial cancer; LVSI Lymphovascular space invasion; NSMP No specific molecular profile; POLEmut Pathogenic polymerase epsilon mutation; MMRd Mismatch repair deficient; p53abn p53 abnormal; TLS Tertiary lymphoid structure.