Table 5. Multivariate analyses of pre- and post-operative parameters and IL-8 and OPN staining scores to predict biochemical recurrence within 24 months.

Cox proportional hazard analyses were performed to identify those parameters that associated with PSA biochemical recurrence within 24 months. The analyses were carried out as follows: 1. only the pre-operative and post-operative parameters; 2. Pre-and post-operative parameters and IL-8 staining score (IL-8 whole model); 3. Pre- and post-operative parameters and OPN staining score (OPN whole model); 4. Pre- and post-operative parameters and OPN and IL-8 staining scores, included as individual parameters (OPN and IL-8 whole model). The analysis for each of the above four categories was also performed by including Gleason sum as a categorical variable (Gleason sum ≥ 8). The significant parameters (P < 0.05) selected by the model are shown. In all of the analyses, IL-8, OPN or IL-8 and OPN staining scores were included as continuous variables.

Parameter	Chi-square	P value	Risk Ratio	95% CI
No IL-8 and/or OPN (Whole model; Gleason continuous or categorized) No pre- or post-operative parameter reached significance
IL-8 (Whole model; Gleason continuous or categorized) No parameter reached significance
OPN (Whole model; Gleason continuous)
OPN	4.32	0.038	1.01*	1.00 – 1.01
OPN (Whole model; Gleason categorized)
OPN	4.16	0.041	1.01*	1.00 – 1.01
IL-8 and OPN (Whole model; Gleason continuous)
OPN	7.13	0.008	1.01*	1.00 – 1.02
IL-8	5.63	0.018	1.01*	1.00 – 1.01
SVI	4.64	0.031	1.74	1.05 – 2.86
IL-8 and OPN (Whole model; Gleason categorized)
OPN	6.59	0.010	1.01*	1.00 – 1.01
IL-8	5.63	0.018	1.01*	1.00 – 1.01
SVI	5.15	0.023	1.78	1.08 – 2.88

^*Increase in risk per unit increase in the staining score.