Skip to main content

Table 2 Cox proportional Hazards models for estimating the effect on time to recurrence

From: Effects of cyclin D1 gene amplification and protein expression on time to recurrence in postmenopausal breast cancer patients treated with anastrozole or tamoxifen: a TransATAC study

  Univariate    Multivariate   
  HR (95% CI) χ12 Pvalue HR (95% CI) χ12 Pvalue
All patients       
   CCND1, amplified vs. nonamplified 2.04 (1.37 to 3.03) 10.51 < 0.001 1.61 (1.08 to 2.41) 4.86 0.030
   Cyclin D1 cytoplasmic intensity, high vs. remainder 0.71 (0.51 to 0.98) 4.43 0.039 0.78 (0.56 to 1.09) 2.15 0.143
   Cyclin D1 nuclear intensity, high vs. remainder 0.91 (0.64 to 1.28) 0.30 0.588 0.85 (0.59 to 1.21) 0.84 0.360
   Cyclin D1 nuclear fraction, 1 to 100% vs. < 1% 0.58 (0.38 to 0.90) 5.28 0.014 0.60 (0.39 to 0.92) 4.75 0.030
Nonamplified patients       
   Cyclin D1 cytoplasmic intensity, high vs. remainder 0.57 (0.39 to 0.82) 9.57 0.003 0.64 (0.44 to 0.95) 5.21 0.022
   Cyclin D1 nuclear intensity, high vs. remainder 0.88 (0.60 to 1.32) 0.38 0.545 0.84 (0.55 to 1.26) 0.76 0.382
   Cyclin D1 nuclear fraction, 1 to 100% vs. < 1% 0.52 (0.34 to 0.81) 7.29 0.004 0.54 (0.35 to 0.85) 6.36 0.012
  1. The χ12 value is based on the likelihood ratio test with the associated P values. Multivariate analyses were adjusted for tumour grade, tumour size, nodal status and Ki67. CI, confidence interval; HR, hazard ratio.