Bioinformatics analysis reveals immune prognostic markers for overall survival of colorectal cancer patients: a novel machine learning survival predictive system

Table 3 Results of cox regression analyses

Variable	Univariate analysis		P value	Multivariate analysis			P value
Variable	HR	95% CI	P value	Coefficient	HR	95% CI	P value
Model cohort (n = 556)
Gender (male/female)	1.323	0.987–1.774	0.061	0.276	1.318	0.976–1.779	0.071
Age (high/low)	1.023	1.011–1.036	< 0.001	0.031	1.032	1.019–1.045	< 0.001
Stage (3–4/1–2)	1.824	1.364–2.439	< 0.001	0.457	1.580	0.877–2.846	0.128
AJCC PM (1–2/0)	4.851	3.389–6.945	< 0.001	1.353	3.870	2.567–5.832	< 0.001
AJCC PN (1–4/0)	1.456	1.091–1.943	0.011	− 0.244	0.784	0.454–1.352	0.381
AJCC PT (3–4/1–2)	2.036	1.042–3.979	0.037	0.457	1.579	0.805–3.100	0.184
Prognostic model (high/low)	4.921	3.472–6.974	< 0.001	1.467	4.337	3.046–6.174	< 0.001
Validation cohort (n = 428)
Gender (male/female)	1.125	0.755–1.678	0.562	− 0.133	0.875	0.583–1.313	0.520
Age (high/low)	1.018	1.002–1.035	0.031	0.034	1.035	1.018–1.052	< 0.001
Stage (3–4/1–2)	2.780	1.842–4.197	< 0.001	0.542	1.719	0.531–5.565	0.366
AJCC PM (1–2/0)	3.073	2.057–4.593	< 0.001	0.773	2.166	1.403–3.343	< 0.001
AJCC PN (1–4/0)	2.688	1.785–4.047	< 0.001	0.448	1.565	0.511–4.799	0.433
AJCC PT (3–4/1–2)	3.299	1.442–7.550	0.005	0.949	2.582	1.082–6.163	0.033
Prognostic model (high/low)	6.787	4.212–10.940	< 0.001	2.092	8.101	4.945–13.270	< 0.001

The median of Prognostic model scores was used as the cut-off value to stratify gastric cancer patients into high risk group and low risk group
AJCC, the American Joint Committee on Cancer; HR, hazard ratio; CI, confidence interval

ISSN: 1471-2105