Skip to main content

Table 3 Univariate and multivariate Cox regression analyses, prognosis-related clinical factors, and clinical independence

From: Co-expression network analysis identifies a gene signature as a predictive biomarker for energy metabolism in osteosarcoma

Variables

Univariate analysis

Multivariable analysis

HR

95% CI of HR

P value

HR

95% CI of HR

P value

TARGET training datasets

 7-gene risk score

  Risk score (high/low)

19.87

4.67–84.42

3.11E−04

28.89

6.25–133.4

1.63E−05

  Age

0.99

0.90–1.09

0.83

1.12

1.00–1.24

0.048

  Gender (male/female)

0.57

0.25–1.27

0.17

0.57

0.24–1.33

0.19

  Metastatic vs non-metastatic

4.18

1.92–9.10

0.0003

5.34

2.21–12.95

0.0002

TARGET datasets

 7-gene risk score

  Risk score (high/low)

17.29

4.10–72.96

1.04E−04

21.07

4.87–91.21

4.54E−05

  Age

0.99

0.91–1.07

0.81

1.08

0.98––1.19

0.098

  Gender (male/female)

0.76

0.36–1.60

0.47

0.76

0.35–1.64

0.48

  Metastatic vs Non-metastatic

4.74

2.27–9.89

3.42E−05

5.69

2.54–12.76

2.36E−05

GSE21257 validation datasets

 7-gene risk score

  Risk score (high/low)

2.84

1.23–6.55

0.014

2.76

1.18–6.46

0.0186

  Age

1.0007

0.99–1.003

0.602

1

0.99–1.003

0.787

  Gender (male/female)

1.4

0.58–3.34

0.44

1.39

0.56–3.41

0.468