TABLE 2

Hazard Ratios, 95% CIs, and P Values from Cox Regression Models on THTF

		Univariate^*		Multivariate^†		Reduced multivariate^‡		Multivariate adjusting for clinical variables^¶
Variable	Range	HR	P	HR	P	HR	P	HR	P
Baseline sum of SUV_max	0–181	1.03 (1.01, 1.04)	<0.001	1.03 (0.99, 1.06)	0.084	1.01 (1.002, 1.03)	0.024	1.02 (1.01, 1.03)	0.004
Baseline maximum SUV_max	0–21.9	1.11 (1.05, 1.17)	0.002	1.12 (0.93, 1.33)	0.26
Baseline average SUV_max	0–18.1	1.15 (1.04, 1.26)	0.012	0.77 (0.51, 1.16)	0.21
Baseline disease sites (n)	0–10¹	1.23 (1.13, 1.33)	<0.001	1.14 (1.02, 1.27)	0.017	1.18 (1.08, 1.29)	<0.001	1.16 (1.03, 1.29)	0.011

↵* Analysis of association between PET variables and THTF.
↵† Includes all 4 PET variables as independent variables.
↵‡ Includes only PET variables that are associated with THTF at P ≤ 0.05 after controlling for each other.
↵¶ Reports association between SUV_max or number of baseline disease sites with THTF after adjusting for clinical variables including age, baseline Gleason score, prostate-specific antigen level, and alkaline phosphatase level.
HR = hazard ratio.
Data in parentheses are 95% CIs.