Skip to main content
. 2022 Jul 13;12(2):1319–1329. doi: 10.1002/cam4.5008

TABLE 3.

Predictors of risk of recurrence

Covariates N Univariate Multivariable a
HR (95% CI) p HR (95% CI) p
Age b 142 1.0 (1.0–1.1) 0.840
Female vs. male 142 2.2 (1.0–5.2) 0.053
Location (colon vs. rectum) 142 2.5 (1.3–5.0) 0.008 4.2 (1.7–10.6) 0.002
Differentiation (well or moderate vs. poor) 124 1.8 (0.7–4.2) 0.174
T‐stage (T1‐2 vs. T3‐4) 142 3.8 (1.5–12.8) 0.013 2.8 (0.8–13.5) 0.152
N‐stage (N0 vs. N1‐2) 142 3.5 (1.8–7.0) <0.001 1.4 (0.5–4.0) 0.547
Intra‐ and/or extra‐mural vascular invasion (no/yes) 113 3.3 (1.4–7.1) 0.004 0.7 (0.2–2.3) 0.589
Lymphovascular invasion (no/yes) 129 3.7 (1.8–7.8) <0.001 1.6 (0.6–4.6) 0.326
Perineural invasion (no/yes) 129 3.5 (1.5–7.4) 0.002 2.6 (0.8–7.4) 0.087
Treatment included surgery (no/yes) 142 2.5 (1.3–5.3) 0.010 0.7 (0.1–4.2) 0.786
Methylated BCAT1/IKZF1 detected (no/yes) 142 3.1 (1.4–6.5) 0.004 2.6 (0.9–6.5) 0.049

Abbreviation: HR, hazard ratio.

a

The multivariable cox proportional hazard regression modeling only included co‐variates identified to be significant in the univariate analysis and cases with missing data were omitted from multivariable analysis (n = 13). In the resulting sub‐population (n = 108), 24 had recurrence.

b

Analyzed as continuous covariate.