Table 1 Univariable and multivariable conditional logistic regression models comparing patients developing metastatic disease with patients free from disseminating disease in a nested case–control study

From: Spatially and functionally distinct subclasses of breast cancer-associated fibroblasts revealed by single cell RNA sequencing

Variablea

n

Univariate models

Multivariate models

HRb

95% CI

P

HRb

95% CI

P

vCAF metagene

 

1.47

1.23–1.76

<0.001

1.66

1.33–2.08

<0.001

mCAF metagene

 

1.28

1.07–1.53

0.005

1.32

1.05–1.66

0.015

Lymph node status

   

<0.001

  

0.003

    Negative

304

1 (ref.)

  

1 (ref.)

  

    Positive

442

2.52

1.69–3.77

 

2.06

1.34–3.16

 

    Unknown

22

1.11

0.36–3.41

 

1.84

0.52–6.46

 

Tumor size, mm

   

0.007

  

0.010

    ≤20

354

1 (ref.)

  

1 (ref.)

  

    >20

398

1.73

1.22–2.44

 

1.81

1.21–2.71

 

    Unknown

16

0.98

0.27–3.59

 

0.78

0.19–3.29

 

Histologic grade

   

<0.001

  

0.012

    Grade 1

68

1 (ref.)

  

1 (ref.)

  

    Grade 2

327

4.86

1.91–12.39

 

3.76

1.38–10.20

 

    Grade 3

328

3.94

1.52–10.23

 

2.51

0.87– 7.20

 

    Unknown

45

3.46

1.09–10.98

 

3.07

0.84–11.20

 

HER2 status

   

<0.001

  

0.001

    Negative

519

1 (ref.)

  

1 (ref.)

  

    Positive

145

2.60

1.74–3.88

 

2.24

1.44–3.49

 

    Unknown

104

0.75

0.44–1.31

 

0.98

0.50–1.92

 

Proliferation metagenec

 

1.20

0.99–1.46

0.061

1.77

1.31–2.40

<0.001

  1. Controls randomly matched to cases by age (<45, 45–55, 55+), adjuvant systemic therapy (endocrine therapy (ET) only, chemotherapy (CT) only, ET + CT), and calendar period of diagnosis (1997–2000, 2001–2005)
  2. aNumerical variables are centered and scaled (standard deviation set to one) in the models
  3. bFor numerical variables, HR is the relative hazard when increasing the variable one standard deviation
  4. cPAM50 proliferation index33, average expression of 11 proliferation genes in the PAM50 gene set
Back to article page