Skip to main content

Table 3 Univariate analysis for each of the 18 markers included in the full model of Figure 1

From: Benefits of biomarker selection and clinico-pathological covariate inclusion in breast cancer prognostic models

Variable

95% CI

Pvalue

[Pathology]

Tumor size

1.064

1.200

0.0003

Age <50 years

0.959

2.075

0.0719

Nuclear grade

1.017

1.676

0.0353

Nodal status

1.717

3.456

2.3 × 10-7

ER (IHC)

0.709

0.926

0.0017

PR (IHC)

0.751

0.986

0.0281

HER2 (IHC)

1.018

1.379

0.0337

[AQUA]

AURKA

1.008

1.019

0.0001

BAG1

0.986

1.008

0.5826

BCL2

0.988

0.999

0.0084

BIRC5

0.999

1.023

0.0642

CCNB1

0.968

1.060

0.5854

CD68

0.975

1.017

0.6866

GRB7

1.006

1.019

0.0008

GSTM1

0.996

1.030

0.1412

KI67

0.994

1.041

0.1722

MMP11

0.996

1.018

0.2349

MYBL2

0.984

1.016

0.9950

  1. For each marker a univariate Cox proportional hazards model is fit to the data using the entire cohort. The 95% confidence interval (CI) is shown together with the log-likelihood test P value. Estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor (HER) 2 were measured by immunohistochemistry (IHC).