Table 3.

Univariate and multivariable Cox proportional hazards regression to evaluate the association between different definitions of PSA dynamics and prostate-cancer–specific mortality. Hazard ratios are shown for a standard deviation increase in the PSA dynamics. Bold face represents significant associations.

	Number of patients	Number of deaths from prostate cancer	Univariate Analysis
			Hazard Ratio	95% CI	P Value	Multivariable P Value*
PSA	594	119	1.23	1.15, 1.31	<0.001	--
Doubling Time
MSKCC26	565	112	0.95	0.80, 1.14	0.6	0.6
Shulman18	561	110	0.97	0.81, 1.17	0.8	0.8
MD Anderson 216	553	111	1.01	0.83, 1.23	0.9	0.9
MD Anderson 116	545	108	0.96	0.80, 1.17	0.7	0.7
Ward21	350	65	1.02	0.83, 1.27	0.8	0.8
Freedland14	327	54	0.83	0.61, 1.12	0.2	0.3
Egawa13	210	34	0.74	0.53, 1.05	0.09	0.055
Trapasso22	194	34	0.88	0.60, 1.30	0.5	0.7
Stephenson20	135	26	0.92	0.57, 1.47	0.7	0.7
Hanks15	112	20	0.81	0.54, 1.22	0.3	0.4
Smith19	59	10	0.72	0.39, 1.34	0.3	0.4
D’Amico12	40	10	0.69	0.27, 1.74	0.4	0.16
Velocity
MSKCC26	594	119	1.02	0.85, 1.21	0.8	0.5
Thompson25	589	118	1.02	0.85, 1.21	0.8	0.5
D’Amico A4	492	106	0.95	0.77, 1.18	0.7	0.01
D’Amico A >2 ng/mL/yr4	492	106	2.62	1.77, 3.88	<0.001	<0.001
Rozhansky23	369	91	0.95	0.76, 1.19	0.7	0.04
Sengupta17	327	54	1.60	1.18, 2.16	0.002	0.3
D’Amico B5	90	15	3.35	2.10, 5.35	<0.001	0.02
D’Amico B >2 ng/mL/yr5	90	15	17.2	5.80, 51.2	<0.001	0.045
Thiel24	82	12	1.51	0.97, 2.35	0.07	0.3
Smith19	59	10	1.26	0.70, 2.27	0.4	0.7

(* Adjusted for PSA)