Skip to main content
. 2016 Aug 8;7(37):59441–59457. doi: 10.18632/oncotarget.11118

Table 2. Univariate and multivariate cox analyses of PFS and OS in the 426 patients with metastatic colorectal cancer (389 with progression, 285 deaths) included in the NORDIC VII study according to pre-treatment plasma TIMP-1 and clinical parameters.

Progression-free Survival Overall Survival
Univariate Cox analyses Multivariate Cox analyses Univariate Cox analyses Multivariate Cox analyses
HR 95% CI p-value HR 95% CI p-value HR 95% CI p-value HR 95% CI p-value
Cetuximab Yes vs No 1.05 0.85–1.29 0.68 0.501 0.27–0.91 b 1.09 0.85–1.40 0.48 0.481 0.25–0.93 a
1.142 0.80–1.63 1.082 0.72–1.63
1.333 0.81–2.19 2.163 1.07–4.33
1.274 0.87–1.84 0.834 0.52–1.03
Plasma TIMP-1 (log)* 1.22 1.07–1.39 0.003 0.845 0.63–1.12 b 1.55 1.33–1.80 < 0.0001 1.045 0.76–1.42 a
1.046 0.80–1.34 1.596 1.17–2.15
2.157 1.04–4.42 4.457 1.73–11.48
1.158 0.81–1.62 1.228 0.83–1.79
Age per 10 years 0.94 0.84–1.04 0.23 0.92 0.82–1.03 0.14 0.97 0.86–1.09 0.60 1.00 0.88–1.14 0.97
Gender, Female vs. Male 1.08 0.88–1.32 0.46 0.92 0.74–1.13 0.41 0.94 0.74–1.19 0.61 1.16 0.90–1.49 0.26
BRAF Mutant vs. WT 2.00 1.41–2.82 < 0.0001 1.73 1.29–2.33 < 0.0001 3.31 2.30–4.77 < 0.0001 4.74 3.10–7.23 < 0.0001
KRAS Mutant vs WT 1.12 0.90–1.39 0.30 1.049 0.74–1.46 b 0.98 0.76–1.27 0.88 1.149 0.78–1.66 a
2.3710 1.29–4.37 2.5610 1.30–5.06
1.3411 0.97–1.85 1.7611 1.16–2.66
1.2712 0.74–2.20 0.6712 0.33–1.41
Metastatic sites > 1 vs 1 1.41 1.13–1.76 0.0026 1.43 1.13–1.81 0.0034 1.59 1.22–2.08 0.0006 1.60 1.20–2.14 0.0015
WHO PS ≥ 1 vs 0 1.61 1.30–1.99 < 0.0001 1.33 1.05–1.68 0.020 1.88 1.47–2.39 < 0.0001 1.52 1.16–1.98 0.0023
Serum CRP, Elevated vs. normal 1.50 1.22–1.85 0.0002 1.44 1.12–1.86 0.0046 1.61 1.25–2.05 0.0002 1.27 0.94–1.71 0.11
Serum CEA, Elevated vs. normal 1.63 1.24–2.14 0.0004 1.39 1.03–1.88 0.034 1.93 1.38–2.71 0.0001 1.70 1.19–2.44 0.0038

HR = Hazard ratio. CI = Confdence interval.

*

Plasma TIMP-1 was included as a log transformed continuous variable (log base 2).

a

P = 0.004 for the interaction Treatment X KRAS X TIMP-1.

b

P = 0.096 for the interaction Treatment X KRAS X TIMP-1.

1

HR for Cetuximab vs no cetuximab for KRAS mutant and TIMP-1 level at the 3rd quartile (409 ng/ml).

2

HR for Cetuximab vs no cetuximab for KRAS WT and TIMP-1 level at the 3rd quartile (409 ng/ml).

3

HR for Cetuximab vs no cetuximab for KRAS mutant and TIMP-1 level at the 1st quartile (201 ng/ml).

4

HR for Cetuximab vs no cetuximab for KRAS WT and TIMP-1 level at the 1st quartile (201 ng/ml).

5

HR for 2-fold difference TIMP-1 levels for mutant KRAS treated with Cetuximab.

6

HR for 2-fold difference TIMP-1 levels for WT KRAS treated with Cetuximab.

7

HR for 2-fold difference TIMP-1 levels for mutant KRAS not treated with Cetuximab.

8

HR for 2-fold difference TIMP-1 levels for WT not treated with Cetuximab.

9

HR for KRAS mutant vs WT receiving cetuximab and TIMP-1 level at the 3rd quartile (409 ng/ml).

10

HR for KRAS mutant vs WT not receiving cetuximab and TIMP-1 level at the 3rd quartile (409 ng/ml).

11

HR for KRAS mutant vs WT receiving cetuximab and TIMP-1 level at the 1st quartile (201 ng/ml).

12

HR for KRAS mutant vs WT not receiving cetuximab and TIMP-1 level at the 1st quartile (201 ng/ml).

For OS, a signifcant interaction between KRAS, TIMP-1 level and treatment with Cetuximab is demonstrated, P = 0.006 in a multivariable model, for PFS the same analysis gives p = 0.078, although not signifcant, the analysis is retained.