Table 3 Univariate and multivariate analyses of predictors of DFS.

From: CT-based radiomics and deep learning models for predicting thyroid cartilage invasion and patient prognosis in laryngeal carcinoma

Clinical characteristics

Univariate cox regression

Multivariate cox regression

HR

95% CI

p

HR

95% CI

p

Age

1.941

0.612–6.155

0.260

   

Sex

1.016

0.990–1.034

0.236

   

Smoking status

1.052

0.457–2.421

0.905

   

Alcohol consumption

1.006

0.637–1.589

0.980

   

Tumor location

Glottic

Ref

     

Supraglottic

1.517

0.934–2.463

0.092

   

Subglottic

0.938

0.285–3.083

0.916

   

CT-reported AC invasion

1.205

0.723–2.010

0.475

   

Histological grade

1.250

0.676–2.311

0.477

   

Clinical T stage

cT1

Ref

  

Ref

  

cT2

0.868

0.425–1.775

0.699

0.606

0.149–2.466

0.484

cT3

2.134

1.169–3.894

0.014*

1.015

0.221–4.664

0.985

cT4

0.750

0.312–1.806

0.522

0.324

0.059–1.780

0.195

Clinical N stage

2.138

1.384–3.303

0.001*

2.191

1.241–3.869

0.007*

Overall clinical stage

I

Ref

  

Ref

  

II

0.999

0.478–2.087

0.998

1.206

0.293–4.964

0.795

III

2.112

1.111–4.015

0.023*

1.284

0.263–6.272

0.757

IV

1.593

0.801–3.167

0.184

0.888

0.155–5.074

0.894

2D DL signature

3.961

1.769–8.867

0.001*

4.666

1.861–11.698

0.001*

  1. DFS disease-free survival, HR hazard ratio, CI confidence interval, AC anterior commissure, DL deep learning.
  2. *p < 0.05.