Table 2.

Cox regression analysis. Univariate and multivariate Cox regression analysis for time to distant metastases in 255 patients.

	Univariate analysis		Multivariate analysis#
	Hazard ratio (95%CI)	p	Hazard ratio (95%CI)	p
Histological grade (1 vs. 2 vs. 3)	3.14 (1.37–7.17)	0.007	0.94 (0.36–2.42)	0.9
Tumor size (≤ 20 mm vs. ≥ 20 mm)	2.18 (1.27–3.75)	0.005	1.58(0.83–3)	0.2
Nodal status (positive vs. negative)	1.62 (0.95–2.79)	0.08	1.30 (0.71–2.37)	0.4
ER high vs. low expression	0.86 (1.18–0.522)	0.5	0.97 (0.56–1.7)	0.9
PgR high vs. low expression	0.42 (0.25–0.7)	0.0007	0.49 (0.26–0.9)	0.02
HER2 high vs. low expression	0.88 (0.55–1.42)	0.6	0.66 (0.37–1.18)	0.2
13 cluster gene classifier*	3.86 (2.32–6.41)	<0.0001	3.26 (1.76–6.05)	0.0002

#Multivariate model contained included 210 patients due to missing values, stratified by population.

*Binary classification using leave-one-out cross-validation.

**Age was not included in the model as 92% of patients were ≥ 50 years of age.

ER: estrogen receptor status represented by ESR1 Affymetrix probe set 205225_at.

PgR: progesterone receptor status represented by PGR Affymetrix probe set 208305_at.

HER2: represented by ERBB2 Affymetrix probe set 216836_s_at.

For ER, PgR and HER2, high vs. low expression groups was defined by generating groups at the median value.