Skip to main content
. 2020 Oct 27;11(43):3807–3817. doi: 10.18632/oncotarget.27774

Table 2. Univariate and multivariate analyses based on Cox’s Proportional Hazards Model for disease-specific survival according to the clinicopathological features evaluated before surgery in stage III colorectal cancer patients.

Variables No. of cases Univariate Analysis Multivariate Analysis
Hazard Ratio 95% Confidence Interval P value Hazard Ratio 95% Confidence Interval P value
Sex
Male/Female 289/223 1.02 0.67-1.55 0.93
Age
≥ 65/< 65 282/230 1.36 0.89–2.09 0.15
Location
Left side/Right side 400/112 1.26 0.76–2.21 0.38
Serum CEA level before surgery (μg/L)
> 5.3/≤ 5.3 178/334 1.67 1.10–2.52 0.017 Not selected
Serum CA19-9 level before surgery (U/ml)
> 37/≤ 37 91/421 1.66 1.00–2.63 0.049 Not selected
Macroscopic type
Type3, 4, 5/Type0, 1, 2 51/461 2.24 1.26–3.73 0.0076 1.82 1.00–3.09 0.047
Maximum diameter of tumor (mm)
≥ 50/< 50 27/275 1.46 0.97–2.22 0.071 Not selected
Tumor intestinal annular propotion
≥ 50/< 50 376/136 2.16 1.25–4.08 0.0049 Not selected
Histological grading
G3/G1, 2 44/468 2.00 1.06–3.48 0.034 Not selected
Depth of tumor
≥ T3/≤ T2 451/61 4.37 1.64–17.79 0.0013 Not selected
Mismatch repair protein
Deficient/Proficient 36/476 1.06 0.45-2.14 0.88
Mesothelin expression
Positive/Negative 61/451 2.32 1.36–3.78 0.0028 2.31 1.34–3.79 0.0033