Table 3

Univariate and multivariate analyses of gene-expression prognostic factors (as continuous variables) for locoregional recurrences

Gene expression by quantitative rt-PCR (continuous variables)

Univariate analysis

Multivariate analysis



HR (RR)

[95% CI]

P

HR (RR)

[95% CI]

P


Estrogen receptor-α

1.22 [0.90–1.66]

0.15

Estrogen receptor-β

0.61 [0.39–0.95]

0.04

0.54

Progesterone receptor

1.13 [0.88–1.45]

0.32

HER1

0.65 [0.37–1.13]

0.10

0.72

HER2

0.89 [0.64–1.25]

0.49

HER3

1.06 [0.55–2.02]

0.87

HER4

1.22 [0.91–1.63]

0.14

MKI 67

0.89 [0.61–1.30]

0.56

Cyclin E1

0.81 [0.48–1.36]

0.43

Cyclin E2

0.75 [0.49–1.14]

0.18

GATA 3

1.61 [0.95–2.73]

0.04

1.49 [0.92–2.39]

0.06

IGF1R

1.41 [0.93–2.14]

0.13

NCoR

0.90 [0.38–2.15]

0.82

NCoA3/AIB1

0.51 [0.21–1.22]

0.14

CYP19 (aromatase)

0.48 [0.28–0.80]

0.003

0.49 [0.29–0.82]

0.003

Skp2

0.52 [0.23–1.17]

0.10

0.18

CAF-1 p150

1.00 [0.48–2.12]

1.00

CAF-1 p60

1.00 [0.97–1.03]

0.94


RR = relative risk for locoregional recurrence calculated with the nonparametric Cox model; CI = confidence interval; HER = human epidermal receptor. Because of lack of prognostic values of the clinical variables in univariate analysis, none was entered in the multivariate analysis.

Bollet et al. Breast Cancer Research 2009 11:R54   doi:10.1186/bcr2343

Open Data