Table 3. Univariate Cox proportional hazards analysis on biochemical recurrence-free survival in the training (n = 410, Probe cg19724470) and validation cohort (n = 259) of prostate cancer patients treated by radical prostatectomy.

	Univariate Cox Analysis
	Training Cohort		Validation Cohort
	Hazard ratio [95% CI]	p-value	Hazard ratio [95% CI]	p-value
pT category (pT3 & pT4 vs. pT2 & pT1)	5.37 [2.14-13.5]	<0.001	2.70 [1.56-4.69]	<0.001
ISUP grading group (1 vs. 2 vs. 3 vs. 4 vs. 5)	1.69 [1.34-2.13]	<0.001	2.00 [1.65-2.44]	<0.001
Surgical margin (R1 vs. R0)	1.49 [0.87-2.56]	0.15	2.31 [1.32-4.06]	0.003
Nodal status (pN1 vs. pN0)	1.84 [1.00-3.35]	0.049	1.39 [0.55-3.50]	0.49
Preoperative PSA level (continuous)	1.04 [1.02-1.05]	<0.001	1.01 [1.00-1.02]	0.11
AR expression§ (AR positive vs. AR negative)	0.96 [0.49-1.89]	0.91	0.76 [0.50-1.16]	0.20
ERG# (ERG-positive vs. ERG-negative)	0.80 [0.40-1.57]	0.51	0.77 [0.40-1.50]	0.44
Age (continuous)	1.02 [0.98-1.06]	0.39	1.01 [0.97-1.06]	0.56
mPD-L1 (continuous)	1.02 [1.00-1.03]	0.024	1.24 [1.08-1.43]	0.002
mPD-L1 (optimized cut-off, mPD-L1high vs. mPD-L1low)	2.60 [1.50-4.51]	0.001	1.90 [1.09-3.31]	0.023

^#Training cohort: ERG fusion (as adopted from [34]); validation cohort: nuclear ERG protein expression (as previously published in [35])

^§Training cohort: AR activity score (as adopted from [34]); validation cohort: AR protein expression (as previously published in [35])